Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Flame ignition studies of conventional and alternative jet fuels and surrogate components
(USC Thesis Other)
Flame ignition studies of conventional and alternative jet fuels and surrogate components
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
FLAME IGNITION STUDIES OF CONVENTIONAL AND ALTERNATIVE JET
FUELS AND SURROGATE COMPONENTS
by
Ning Liu
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements of the Degree
DOCTOR OF PHILOSOPHY
(MECHANICAL ENGINEERING)
May 2013
Copyright 2013 Ning Liu
ii
Dedication
This dissertation is dedicated to my loving husband, Geng Fu, my great parents and
parents-in-law, Xingyun Liu, Meichen Li, Guobin Fu and Chunxiang Feng, for their
endless love, support, and encouragement to reach my dreams.
iii
Acknowledgements
I would first like to express my sincere gratitude to my advisor, Professor Fokion N.
Egolfopoulos for his great support, encouragement, and constructive guidance throughout
my research to pursue my Ph.D. degree. Prof. Egolfopoulos opened my eyes to the world
of combustion science and imparted me the fundamental knowledge that will definitely
benefit me in my future career. The genuine enthusiasm and desire for research that
radiated from him motivated me. He taught me to reach big goals with scientific
carefulness and to be brave in life. I will forever be grateful for everything he taught me.
I would also like to thank to my doctoral committee members, Professor Theodore T.
Tsotsis, Professor Paul Ronney, Professor Hai Wang, and Professor Satwindar S. Sadhal,
for their valuable efforts and suggestions on my research. Their wisdom and insightful
criticism inspired me to achieve a high standard. Thanks as well to the staff of the
Department of Aerospace and Mechanical Engineering, Ms. Silvana Martinez-Vargas,
Ms. Samantha Graves, Ms. April Mundy, and Mr. Jorge Castilla, for all of the
professional support and daily assistance during the past five years.
I wish to express thanks to all my colleagues for sharing their experience and
discussions. I would like to especially thank Dr. Adam T. Holley and Dr. Chunsheng Ji.
Adam offered his academic help and research advice in the first year of my graduate
study. Chunsheng shared his brilliant ideas and experimental knowledge, and provided
necessary help when I was first learning the numerical codes in combustion study.
Most importantly, this work could not have been possibly accomplished without the
encouragement and patience from my loving husband who is my inspiration of this
iv
dissertation. I have never felt lonely for having him along the journey. He is an
absolutely invaluable treasure of my life. To my parents, thank you for always being
there for me and I am grateful that you unconditionally love and believe in me throughout
my life. It’s my big fortune to have my dear parents-in-law and I would like to thank
them for their continuous support, understanding, love, and absolute trust over the years.
Finally, I wish to acknowledge the financial support sponsored by the U.S. Air Force
Office of Scientific Research AFOSR to complete this work.
v
Table of Contents
Dedication
ii
Acknowledgements
iii
List of Tables
viii
List of Figures
ix
Abstract
xvii
Chapter 1: Introduction 1
1.1 Overview and Significance 1
1.1.1 Jet Fuels and their Surrogates 1
1.1.2 Classification and Application of Ignition 4
1.1.3 Criterion and Kernel of Ignition 5
1.2 Motivation and Objectives 9
1.3 Organization of Dissertation 11
1.4 References 13
Chapter 2: Experimental Approach 16
2.1 Experimental Apparatus 16
2.2 Experimental Procedures 19
2.2.1 Determination of Ignition Limits 19
2.2.2 Vaporization of Liquid Fuels 21
2.3 Experimental Uncertainties 22
2.4 References 25
Chapter 3: Numerical Approach 26
3.1 Codes Description 26
3.2 Effects of Local Strain Rate 30
3.3 Kinetic Models 31
3.4 References 31
Chapter 4: Ignition Characteristics of Premixed Alternative Gaseous Fuel
Blends Flames
34
4.1 Introduction 34
4.2 Experimental Approach 36
4.3 Modeling Approach 37
4.4 Results and Discussion 38
4.5 Concluding Remarks 44
4.6 References 44
vi
Chapter 5: Ignition of Non-Premixed C
3
-C
12
n-alkane Flames 47
5.1 Introduction 47
5.2 Experimental Approach 51
5.3 Modeling Approach 52
5.4 Results and Discussion 53
5.4.1 Ignition of n-dodecane Flames 53
5.4.2 Ignition of C
3
-C
10
n-alkane Flames 57
5.4.3 Comparisons of the Ignition of C
3
-C
12
n-alkane Flames 62
5.5 Concluding Remarks 72
5.6 References 73
Chapter 6: Ignition of Non-Premixed Counterflow Flames of Octane and
Decane Isomers
78
6.1 Introduction 78
6.2 Experimental Approach 80
6.3 Modeling Approach 83
6.4 Results and Discussion 85
6.4.1 Ignition of Octane and Decane Isomer Flames 85
6.4.2 Comparison of Ignition Temperatures 88
6.5 Concluding Remarks 100
6.6 References 100
Chapter 7: Ignition of Non-Premixed Cyclohexane and Mono-Alkylated
Cyclohexane Flames
105
7.1 Introduction 105
7.2 Experimental Approach 108
7.3 Modeling Approach 110
7.4 Results and Discussion 111
7.4.1 Ignition of Cyclohexane Flames 111
7.4.2 Comparison of Ignition Propensity of Cyclohexane and n-
hexane Flames
116
7.4.3 Ignition of Methyl-, Ethyl-, n-Propyl-, and n-Butyl-
Cyclohexane Flames
119
7.5 Concluding Remarks 129
7.6 References 130
Chapter 8: Ignition Characteristics of Non-Premixed Binary Fuel Blends
Flames
133
8.1 Introduction 133
8.2 Experimental Approach 136
8.3 Modeling Approach 137
8.4 Results and Discussion 138
vii
8.4.1 Ignition of Non-Premixed Flames of C
2
H
4
and Validation
of Detailed Kinetic Models
138
8.4.2 Ignition of Non-Premixed Flames of n-C
12
H
26
, and n-
C
12
H
26
/C
2
H
4
Blends
144
8.4.3 Effects of Ethylene on the Ignition of n-Dodecane Flames 145
8.5 Concluding Remarks 151
8.6 References 152
Chapter 9: Flame Ignition Studies of Conventional and Alternative Jet Fuels 155
9.1 Introduction 155
9.2 Experimental Approach 157
9.3 Modeling Approach 160
9.4 Results and Discussion 160
9.5 Concluding Remarks 164
9.6 References 164
Chapter 10: Mixing Rules of Ignition Phenomena 167
10.1 Introduction 167
10.2 A Mixing Rule in Homogeneous Systems 168
10.3 A Mixing Rule in Counterflow Flames 171
10.4 Concluding Remarks 177
10.5 References 177
Chapter 11: Conclusions and Recommendations 178
11.1 Conclusions 178
11.2 Recommendations for Future Work 182
Bibliography 184
viii
List of Tables
Table 4.1 Composition of Blended Fuels. 37
Table 6.1 Octane isomers and corresponding boiling temperatures. 82
Table 6.2 Decane isomers and corresponding boiling temperatures. 82
Table 7.1 Cyclohexane, mono-alkylated cyclohexanes and
corresponding boiling temperatures.
109
Table 9.1 Jet fuel properties and detailed composition on a per mass
basis.
159
ix
List of Figures
Figure 1.1 Species profiles of H, OH, CH
3
, HO
2
and temperature profile
of premixed n-dodecane flames at ignition state computed
using JetSurf 1.0. ( = 1.0, separate distance, L = 2.2 cm,
unburned n-dodecane/air mixture temperature, T
u
= 298 K and
velocity, V
u
= 100 cm/s.)
6
Figure 1.2 Species profiles of H, OH, CH
3
, HO
2
and temperature profile
of non-premixed n-dodecane flames at ignition state computed
using JetSurf 1.0. (Fuel mole fraction, X
F
= 10 %, separate
distance, L = 2.2 cm, unburned n-dodecane/N
2
mixture
temperature, T
u
= 298 K and velocity, V
u
= 100 cm/s.)
7
Figure 1.3 Reaction rates and temperature profile of premixed n-
dodecane flames at ignition state computed using JetSurf 1.0.
( = 1.0, separate distance, L = 2.2 cm, unburned n-
dodecane/air mixture temperature, T
u
= 298 K and velocity, V
u
= 100 cm/s.)
8
Figure 1.4 Reaction rates and temperature profile of non-premixed n-
dodecane flames at ignition state computed using JetSurf 1.0.
(Fuel mole fraction, X
F
= 10 %, separate distance, L = 2.2 cm,
unburned n-dodecane/N
2
mixture temperature, T
u
= 298 K and
velocity, V
u
= 100 cm/s.)
9
Figure 2.1 Schematic of the experimental configuration. 17
Figure 2.2 Schematic of the thermocouple. 18
Figure 2.3 Radial temperature profiles measured at different distances, z,
from the upper burner exit; ( ◆) z = 0 mm; ( ■) z = 2 mm; ( ◇)
z = 4 mm; ( ▲) z = 6 mm; ( □) z = 8 mm.
20
Figure 3.1 Schematic of a typical S-curve determined in the opposed-jet
configuration.
27
Figure 3.2 Computed responses of maximum H radical mass fraction to
hot air boundary temperature by using mixture-averaged and
multicomponent formulations for a non-premixed iso-octane
flame, X
F
= 5.10 %, T
u
= 401 K, K = 130 s
-1
.
28
x
Figure 3.3 Axial velocity profile along the stagnation streamline on the
fuel-side.
30
Figure 4.1 Schematic of the experimental configuration. 36
Figure 4.2 Experimental and computed T
ign
’s of premixed flames;
( ◇)( ―) Fuel # 1, 30%CH
4
/60%H
2
/10%CO; ( ♦)(···) Fuel # 2,
5%CH
4
/95%H
2
; ( ■)( ‐‐‐) Fuel # 3a, 10% CH
4
/32%H
2
/58%CO;
( ▲)(– · ‒) Fuel # 4, 25% CH
4
/10%CO
2
/54%H
2
/11%CO.
38
Figure 4.3 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for Fuel # 1 at = 0.40 and = 0.61.
39
Figure 4.4 Concentration profiles of CH
4
and H
2
for Fuel # 1 and Fuel #
3a at = 0.50 at ignition state.
40
Figure 4.5 Concentration profiles of CH
3
and H for Fuel # 1 and Fuel #
3a at = 0.50 at ignition state.
41
Figure 4.6 Experimental and computed T
ign
’s of premixed flames; ( ■)(―)
Fuel # 3a, 10% CH
4
/32%H
2
/58%CO; ( □)(‐‐‐) Fuel # 3b, 10%
C
3
H
8
/32%H
2
/58%CO.
41
Figure 4.7 Reaction path analyses of flames of (a) C
3
H
8
/air, (b) Fuel # 3b
(10 % C
3
H
8
/ 32 % H
2
/ 58 % CO)/air at the ignition state and
= 0.70 using USC Mech II. M represents a third body. The
numbers indicate the conversion percentages.
43
Figure 5.1 Experimental and computed T
ign
of n-C
12
H
26
flames; ( ─)
Simulation results using the JetSurF 1.0 reaction model with
multi-component transport coefficient formulation; ( ┄ )
Simulation results with mixture-averaged transport coefficient
formulation.
54
Figure 5.2 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients, computed for n-C
12
H
26
flames at X
F
= 2% and
X
F
= 10%.
56
Figure 5.3 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for n-C
12
H
26
flames at X
F
= 2% and
X
F
= 10%.
56
xi
Figure 5.4 Experimental and computed T
ign
’s; (a) C
3
H
8
, (b) n-C
5
H
12
, (c)
n-C
6
H
14
, and (d) n-C
7
H
16
flames; Symbols: experimental data;
Lines: simulation results using the JetSurF 1.0 kinetic model
with multi-component transport coefficient formulation.
58
Figure 5.5 Experimental and computed T
ign
’s; (a) n-C
8
H
18
, (b) n-C
9
H
20
,
and (c) n-C
10
H
22
flame; Symbols: experimental data; Lines:
simulation results using the JetSurF 1.0 kinetic model with
multi-component transport coefficient formulation.
59
Figure 5.6 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients, computed for C
3
H
8
flames at X
F
= 3% and
X
F
= 10%.
60
Figure 5.7 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for C
3
H
8
flames at X
F
= 3% and
X
F
= 10%.
61
Figure 5.8 Comparison of experimentally and numerically determined
T
ign
’s; ( ◇ )( ─) n-C
5
H
12
, ( ■)(…) n-C
7
H
16
, and ( ▲)( ┄) n-
C
12
H
26
flames. Symbols: present experimental data; Lines:
simulation results using the JetSurF 1.0 kinetic model with
multi-component transport coefficient formulation.
62
Figure 5.9 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for n-C
5
H
12
, n-C
7
H
16
, and n-C
12
H
26
flames; (a) X
F
= 2%, (b) X
F
= 10%.
64
Figure 5.10 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for n-C
5
H
12
, n-C
7
H
16
, and n-C
12
H
26
flames; (a) X
F
= 2%, (b) X
F
= 10%; the logarithmic sensitivity
coefficient on the main branching reaction is shown for
comparison purposes.
66
Figure 5.11 Concentration profiles of C
2
H
4
, n-C
5
H
12
, and n-C
12
H
26
at (a)
X
F
= 2%, T
air
= 1333 K and (b) X
F
= 10%, T
air
= 1226 K; ( ┄)
n-C
5
H
12
flames; ( ─) n-C
12
H
26
flames.
67
Figure 5.12 Concentration profiles of H, OH, n-C
5
H
12
, and n-C
12
H
26
at (a)
X
F
= 2%, T
air
= 1333 K and (b) X
F
= 10%, T
air
= 1226 K; ( ┄)
n-C
5
H
12
flames; ( ─) n-C
12
H
26
flames.
69
xii
Figure 5.13 Comparison of experimentally determined T
ign
’s; (+) C
3
H
8
, (*)
n-C
5
H
12
, ( ♦) n-C
6
H
14
, ( ■) n-C
7
H
16
, ( □) n-C
8
H
18
, ( ∆) n-C
9
H
20
,
( ◇) n-C
10
H
22
, and ( ▲) n-C
12
H
26
. (a) Data based on fuel mole
fraction, Inset: detail of the region of fuel mole fractions
below 4%; (b) Data based on fuel mass fraction.
71
Figure 6.1 Carbon skeletal structure of 2,7-dimethyloctane (C
10
H
22
-27)
with carbon sites labeled.
83
Figure 6.2 Experimental and computed T
ign
’s of n-octane, iso-octane, 3-
MHP, and 2,5-DMH flames. Symbols: experimental data.
Lines: simulation results.
85
Figure 6.3 Experimental and computed T
ign
’s of n-decane and 2,7-DMO
flames. Symbols: experimental data. Lines: simulation
results.
87
Figure 6.4 Experimental and computed T
ign
’s of C
8
isomers flames. (○)(-
∙-) n-octane, ( ■)(---) 3-MHP, (∆)(∙∙∙) 2,5-DMH and ( ◇)( ─)
iso-octane. Symbols: experimental data; Lines: simulation
results using Model I for n-octane, 3-MHP, and 2,5-DMH,
and Model II(b) for iso-octane.
88
Figure 6.5 Experimental and computed T
ign
’s of C
10
isomers flames.
( ◇)( ─) n-decane and (○)(---) 2,7-DMO. Symbols:
experimental data; Lines: simulation results using Model I.
89
Figure 6.6
Experimental and computed T
ign
’s of 2,5-DMH ( ◇)( ─) and
2,7-DMO (●)(---)(∙∙∙) flames. Symbols: experimental data;
Lines: simulation results using Model I; (∙∙∙) simulations with
diffusivity of 2,7-DMO changed to that of 2,5-DMH.
90
Figure 6.7 Ranked logarithmic sensitivity coefficients of T
ign
on kinetics
computed for a X
F
= 4% 3-MHP flame using Model I.
91
Figure 6.8 Ranked logarithmic sensitivity coefficients of T
ign
on kinetics
computed for a X
F
= 4% 2,5-DMH flame using Model I.
92
Figure 6.9 Ranked logarithmic sensitivity coefficients of T
ign
on kinetics
computed for a X
F
= 4% iso-octane flame using Model II(b).
93
Figure 6.10 Ranked logarithmic sensitivity coefficients of T
ign
on kinetics
computed for a X
F
= 4% 2,7-DMO flame using Model I.
94
xiii
Figure 6.11 Reaction path analysis of flames of 2,5-DMH (a) and iso-
octane (b) at the ignition state and X
F
= 4% using (a) Model I,
(b) Model II(b). The numbers indicate the conversion
percentages.
96
Figure 6.12 Concentration profiles of iC
4
H
8
in X
F
= 4% 3-MHP, 2,5-
DMH, iso-octane and 2,7-DMO flames at the ignition state.
97
Figure 6.13 Concentration profiles of C
2
H
3
and CH
3
in X
F
= 4% n-octane,
3-MHP, 2,5-DMH, iso-octane and 2,7-DMO flames at the
ignition state.
98
Figure 7.1 Experimental and computed T
ign
’s of CHX flames. ( ♦)
experimental data, (─) simulations using Model I, (---)
simulations using Model II.
111
Figure 7.2 Reaction path analysis of a X
F
= 6% CHX flame at the
ignition state using (a) Model I, (b) Model II. The numbers
indicate the conversion percentages.
112
Figure 7.3 Logarithmic sensitivity coefficients of T
ign
on kinetics for a
X
F
= 6% CHX flame, computed using Model I.
114
Figure 7.4 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients for a X
F
= 6% CHX flame, computed using Model
I.
115
Figure 7.5 Experimental and computed T
ign
’s of CHX and n-C
6
H
14
flames. Symbols: experimental data ( ♦) CHX and ( ▲) n-
C
6
H
14
. Lines: simulation results using Model I, (---) CHX and
( ─) n-C
6
H
14
.
116
Figure 7.6 Computed T
ign
of ( ─) CHX and (∙ ‐ ∙) n-C
6
H
14
flames using
Model I with original Lennard-Jones parameters, and (---)
simulations with modified Lennard-Jones parameters of CHX.
117
Figure 7.7 Concentration profiles of (a) H and HCO, (b) C
2
H
3
and CH
3
at
X
F
= 6%, T
air
= 1223 K and K = 120s
-1
, (---)cyclohexane
flames; (—) n-hexane flames.
118
Figure 7.8 Experimental and computed T
ign
’s for methyl-CHX, ethyl-
CHX, n-propyl-CHX, and n-butyl-CHX flames. Symbols:
experimental data. Lines: simulation results using Model I.
121
xiv
Figure 7.9 Comparison of experimentally determined T
ign
’s for flames of
( ♦) CHX, ( □) methyl-CHX, ( ∆) ethyl-CHX, ( ○) n-propyl-
CHX, (
*
) n-butyl-CHX, and ( ▲) n-C
6
H
14
.
122
Figure 7.10 Reaction path analysis of a X
F
= 6% ethyl-CHX flame at the
ignition state using Model I. The numbers indicate the
conversion percentages.
122
Figure 7.11 Reaction rates of initial C-C bond fission computed of non-
premixed CHX, methyl-CHX and n-propyl-CHX flames at
ignition state for X
F
= 6% and K = 120s
-1
, ( ‒·· ‒) CHX; (---)
methyl-CHX; and (—) n-propyl-CHX.
123
Figure 7.12 Logarithmic sensitivity coefficients of T
ign
on kinetics for
X
F
= 6% mono-alkylated CHX flames, computed using Model
I.
124
Figure 7.13 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients for X
F
= 6% mono-alkylated CHX flames,
computed using Model I.
125
Figure 7.14 Concentration profiles of (a) H, (b) HCO, (c) C
2
H
3
, and (d)
CH
3
at X
F
= 6%, T
air
= 1223 K and K = 120s
-1
for non-
premixed mono-alkylated cyclohexane flames.
127
Figure 8.1 Major products of n-dodecane decomposition under 500 K,
800 K, and 1000 K.
134
Figure 8.2 Schematic of the experimental configuration. 136
Figure 8.3 Experimental and computed T
ign
’s of non-premixed C
2
H
4
flames.
140
Figure 8.4 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for C
2
H
4
flames; T
u
= 453 K, K = 130s
-
1
, X
F
= 6%.
141
Figure 8.5 CH
2
O concentration profiles along the center of the stream
line computed for C
2
H
4
flames at T
air
= 1173 K and X
F
= 6%.
142
Figure 8.6 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for C
2
H
4
flames; T
u
= 453 K, K = 130s
-
1
, X
F
= 6%.
143
xv
Figure 8.7 Experimental and computed T
ign
’s of n-C
12
H
26
, and n-
C
12
H
26
/C
2
H
4
blends flames. Symbols: experimental data.
Lines: simulation results.
145
Figure 8.8 Experimental T
ign
’s of C
2
H
4
( ♦), n-C
12
H
26
( ▲), 90% n-C
12
H
26
/ 10% C
2
H
4
( ○), 80% n-C
12
H
26
/ 20% C
2
H
4
( ▪), and 50% n-
C
12
H
26
/ 50% C
2
H
4
( ▲) flames.
146
Figure 8.9 Experimental T
ign
’s of C
2
H
4
, n-C
12
H
26
, 90% n-C
12
H
26
/ 10%
C
2
H
4
, 80% n-C
12
H
26
/ 20% C
2
H
4
, and 50% n-C
12
H
26
/ 50%
C
2
H
4
flames versus mole fractions of C
2
H
4
at X
F
= 3.5 % and
X
F
= 7 %.
147
Figure 8.10 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for n-C
12
H
26
and n-C
12
H
26
/C
2
H
4
blends
flames with X
F
= 6%.
148
Figure 8.11 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for n-C
12
H
26
and n-C
12
H
26
/C
2
H
4
blends
flames with X
F
= 6%.
149
Figure 8.12 C
2
H
4
and n-C
12
H
26
concentration profiles computed for n-
C
12
H
26
and n-C
12
H
26
/C
2
H
4
blends flames at ignition state with
X
F
= 6%. ( ―) n-C
12
H
26
, (···) 90% n-C
12
H
26
/ 10% C
2
H
4
, (– · –
)80% n-C
12
H
26
/ 20% C
2
H
4
, ( ‒ ‒ ‒) 50% n-C
12
H
26
/ 50% C
2
H
4
.
150
Figure 9.1 Comparison of fuel composition distribution for JP-7, JP-8, S-
8, and Shell-GTL on a per mass basis.
158
Figure 9.2 Experimentally determined T
ign
’s as a function of Fuel/N
2
mass ratio at T
u
= 453 K and K = 130 s
-1
of JP-7 ( △), JP-8 ( ◇
), S-8 ( ■), Shell-GTL ( ▲), n-C
10
H
22
( ●), and n-C
12
H
26
/air (
◆) flames.
161
Figure 9.3 Computed H mass fraction response to the hot boundary
temperatures using JetSurF 2.0 for a fuel mole fraction, X
F
= 6
% at T
u
= 453 K and K = 130 s.
-1
( ― ) n-C
12
H
26
, ( --- ) 82.6
% n-C
12
H
26
+ 17.4 % methylcyclohexane per mass basis, ( ∙ ‒
∙ )86.5 % n-C
12
H
26
+ 13.5 % toluene per mass basis.
163
Figure 10.1 Computations using USC_Mech_II and estimations of
ignition delay times of binary fuel blends at = 1. (a) 50 %
CH
4
/ 50 % H
2
, (b) 50 % CH
4
/ 50 % C
3
H
8
, (c) 20 % CH
4
/ 80
% (H
2
+ CO), H
2
: CO = 1 : 1, per volume.
169
xvi
Figure 10.2
Ignition response of non-premixed methane flames. ( ◆ )
100% CH
4
; ( ▲) 75% CH
4
; ( ■) 50% CH
4
; ( ◇) 10% CH
4
in
N
2
. Lines are best fits of computational values.
171
Figure 10.3 Ignition response of non-premixed propane flames. ( ▲) 75%
C
3
H
8
; ( ■) 50% C
3
H
8
; ( ◇) 10% C
3
H
8
in N
2
. Lines are best fits
of computational values.
172
Figure 10.4 Ignition response of non-premixed methane/propane blends
flames. ( ▲) 75% fuel blend; ( ■) 50% fuel blend; ( ◇) 10%
fuel blend in N
2
. Lines are best fits of computational values.
172
Figure 10.5 Computations and estimations of K
ign
of non-premixed flames
of methane/propane blends with volume ratio of 1:1. (a) 75%
fuel blend; (b) 50% fuel blend; (c) 10% fuel blend in N
2
; ( ▲)
computed results; ( ■) estimations. Lines are best fits of
computed and estimated values.
174
Figure 10.6 Computations and estimations of overall activation energy of
non-premixed flames of methane/propane blends with volume
ratio of 1:1 as a function of fuel mole fractions. ( ▲)
computed results; ( ■) estimations.
176
Figure 10.7 Computations and estimations of overall pre-exponential
factor of non-premixed flames of methane/propane blends
with volume ratio of 1:1 as a function of fuel mole fractions.
( ▲) computed results; ( ■) estimations.
176
xvii
Abstract
Practical jet fuels are widely used in air-breathing propulsion, but the chemical
mechanisms that control their combustion are not yet understood. Thousands of
components are contained in conventional and alternative jet fuels, making thus any
effort to model their combustion behavior a daunting task. That has been the motivation
behind the development of surrogate fuels that contain typically a small number of neat
components, whose physical properties and combustion behavior mimic those of the real
jet fuel, and whose kinetics could be modeled with increased degree of confidence.
Towards that end, a large number of experimental data are required both for the real fuels
and the attendant surrogate components that could be used to develop and validate
detailed kinetic models. Those kinetic models could be used then upon reduction to
model a combustor and eventually optimize its performance.
Among all flame phenomena, ignition is rather sensitive to the oxidative and
pyrolytic propensity of the fuel as well as to its diffusivity. The counterflow
configuration is ideal in probing both the fuel reactivity and diffusivity aspects of the
ignition process and it was used in the present work to determine the ignition
temperatures of premixed and non-premixed flames of a variety of fuels relevant to air-
breathing propulsion. The experiments were performed at atmospheric pressure, elevated
unburned fuel mixture temperatures, and various strain rates that were measured locally.
Several recent kinetic models were used in direct numerical simulations of the
experiments and the computed results were tested against the experimental data.
Furthermore, through sensitivity, reaction path, and structure analyses of the computed
xviii
flames, insight was provided into the dominant mechanisms that control ignition. It was
found that ignition is primarily sensitive to fuel diffusion and secondarily sensitive to
chemical kinetics and intermediate species diffusivities under the low fuel concentrations.
As for the detailed high temperature oxidation chemistry, ignition of normal, branched,
and cyclic alkane flames were found to be sensitive largely to H
2
/CO and C
1
-C
4
small
hydrocarbon chemistry, while for branched alkanes fuel-related reactions do have
accountable effect on ignition due to the low rate of initial fuel decomposition that limits
the overall reactions preceding ignition. Analyses of the computed flame structures
revealed that the concentrations of ignition-promoting radicals such as H, HCO, C
2
H
3
,
and OH, and ignition-inhibiting radicals such as C
3
H
6
, aC
3
H
5
, and CH
3
are key to the
occurrence of ignition.
Finally, the ignition characteristics of conventional and alternative jet fuels were
studied and were to correlate with the chemical classifications and diffusivities of the
neat species that are present in the practical fuel.
1
Chapter 1
Introduction
1.1 Overview and Significance
Liquid fuels, such as gasoline, diesel, and jet fuels, are ideal for transportation, cars
and aircrafts, because of their convenient handling and storing, high energy content,
performance, and availability. According to the Annual Energy Outlook 2012 from U.S.
Energy Information Administration, petroleum-based liquid fuels constitute the largest
source of U.S. energy consumption, which is approximately 37 percent of total energy
consumption in 2010; among the total imports of energy 17 percent is liquid fuels [1].
1.1.1 Jet Fuels and their Surrogates
The United States (US) consumed 1.43 million barrels of jet fuels per day in 2011 and
1.41 million barrels per day in the first two quarters in 2012 [1]. The US military, as the
single largest petroleum consumer in the world with a jet fuel consumption of 62 % at
volume basis of all petroleum products [2], has been at the forefront of jet fuels
2
development. The forecasting result reported by Cheze et al. [3] featured that world jet
fuel demand will grow yearly at a rate of 1.9% and increase by about 38% from 2008 to
2025. Such requirement impels the importance of the development of jet fuels to high-
energy efficiency and enhanced operability.
The development of aviation fuel is generally determined by the input from petroleum
refiners, engine performance, aircraft industry, and government regulations. Petroleum-
derived jet fuels, mainly dominated by several basic structural classes of compounds,
such as n-alkanes, iso-alkanes, cyclo-alkanes, and aromatics, are chosen mainly because
of the availability. Aviation gasoline (avgas) was the first “jet fuels” in the United States
since the early development of turbojet engines [4]. Subsequently as jet-powered aircraft
was evolved, avgas could not satisfy the operation demand due to the high volatility of
gasoline causing engine malfunction. The development of jet fuels was promoted by the
operational requirements including fluidity, corrosion protection, heat content, and
stability. In 1951, a mixture of the gasoline and kerosene was issued for Jet Propellant
JP-4 and matured as an operational jet fuel [4]. Following that, according to higher
volumetric heating values and less soot, kerosene-type fuels were established as jet fuels,
such as JP-A and JP-8. However, due to the national concerns about energy security and
petroleum availability, alternative jet fuels are attracting much attention especially with
the recent increase in the price of crude oil.
Alternative jet fuels for aviation have been considered since the early days of turbine
engines that addresses associated economic, environmental, and energy security concerns.
They are produced from the reforming of coal, natural gas and biomass securing from
3
depleting crude oil sources. S-8 is a type of Fischer-Tropsch (F-T) fuels derived from
natural gas by Syntroleum. Another example is an F-T fuel derived from gas-to-liquid
(GTL) by Shell, hereafter referred to as Shell-GTL. Those two fuels are all composed of
essentially n-alkane and weakly branched alkanes [5,6]. F-T fuel is a cleaner burning
fuel than conventional jet fuel because it is free of aromatic compounds and there are no
sulfur dioxide (SO
2
) or sulfuric acid (H
2
SO
4
) aerosol emissions. However, the absence of
aromatic in F-T synthetic fuel will cause the shrinkage of elastomers that used in aircraft
fuel systems to swell, which may lead to fuel leaks, suggesting the blends of F-T fuels
and petroleum-derived jet fuels or to find an additive that would provide the necessary
polymeric swelling behavior in the absence of aromatics. Sasol first examined the
synthetic fuel blends and was been granted approval in 2008 to supply a fully synthetic
product to Johannesburg International Airport [7].
Jet fuels are composed of hundreds and often thousands of neat hydrocarbons and
their compositions vary significantly due to the source of crude oil and the refinery
process, posing challenges to understanding of the combustion characteristics of those
complex fuel blends. To tackle issues related to combustion control, reduction of fuel
consumption, and emissions of pollutants and greenhouse gases in practical combustion
systems, it is essential to have establish a reasonable understanding of the controlling
physical and chemical mechanisms of the combustion process. Due to the chemical
complexity of practical jet fuels their combustion properties cannot be modeled from first
principle and understood. Therefore, it is essential to formulate surrogate fuel blends
containing a limited number of representative hydrocarbon compounds by matching the
4
physical and chemical properties of the practical fuel. The combustion properties of
surrogate fuels can be thus computationally tracktable.
For the past few decades, several studies on the development of fuel surrogates have
been carried out (e.g., [8-17]). In recent efforts to establish appropriate surrogate fuels
additional neat components have been introduced to the surrogates, such as for example
n-butyl-cyclohexane and 2,7-dimethyloctane, which did not get enough attention
previously. Moreover, there still exist notable disagreements among experimental data
and models predictions, which complicates ongoing efforts to formulate reliable
surrogate fuels.
1.1.2 Classification and Application of Ignition
Ignition is a rathe important limit combustion phenomenon, which can be achieved in
two ways [18]. The first one involves heat resulting in thermal ignition. Chemical
reactions respond to heat addition and can generate more heat that increases the gas
temperature. Ignition is attained when the rate of chemical heat generation in the reaction
zone is much faster than heat loss, resulting in thermal runaway and eventually ignition.
The second one involves sufficient amount of radical production that exceeds the loss of
radical, resulting in radical runaway and eventually ignition.
In many practical combustion systems, ignition depends not only on the physical and
chemical properties of fuels but also the mixing and transport processes. The
performance of the systems is related to great extent to ignition, i.e., engine knock in
5
spark-ignition engines, ignition delay in diesel engines, and high-altitude re-light in jet
engines.
The two most common combustion modes are premixed and non-premixed. The
spark-ignition engine operates under the conditions that fuel and oxidizer are mixed at the
molecular level before ignition. In case of diesel engines and jet engines on the other
hand, fuel is sprayed into air so that the reactants are initially unmixed.
Various experimental approaches have been used to investigate the combustion
characteristics of jet fuels and their surrogates, such as shock tubes (e.g., [17, 19-21]),
rapid compression machines (e.g., [22-24]), flow reactors (e.g., [12,25]), as well as
spherical (e.g., [26]) and couterflow flames (e.g., [6,27-30]). Most of the ignition
investigations were done in homogenous systems with focus on chemical kinetics.
However, in jet engines ignition initiates under non-homogenous conditions and in the
presence of notable gradients of temperature, species concentrations, and flow velocities,
which augment the importance of convective and diffusive effects on ignition. The
counterflow configuration was adopted in the present study as it is ideal study the flame
properties at its elementary level and it can additionally be modeled directly using one-
dimensional formulation, allowing thus for the assessment of both chemical kinetics and
convective/diffusive transport of mass and heat.
1.1.3 Criterion and Kernel of Ignition
Theoretically, the response of counterflow non-premixed flames can be presented
typically in the form of S-curve [18] by plotting either the maximum reaction temperature
or
br
in
b
h
m
cr
ra
[3
th
r the burning
ranches, nam
ntensely bur
efore it goe
eat is gener
manner; thus
riterion of ig
By increa
adical pool i
31,32], in wh
his concept h
Figure 1
g rate versus
mely 1) we
rning branch
s to the uns
ated so fast
ignition occ
gnition can b
asing the hot
s built up in
hich it is hig
herein.
1.1 Species
premix
JetSur
dodeca
cm/s.)
s a system D
eakly reactin
h. By incre
stable branch
in the react
curs correspo
be readily de
t boundary t
a narrow sp
ghly chemica
s profiles of
xed n-dodec
rf 1.0. ( = 1
ane/air mixtu
Damkohler nu
ng branch,
easing D
a
al
h a turning p
tion zone an
onding to th
etermined tha
emperature,
patial region.
al reactive.
f H, OH, CH
cane flames
1.0, separate
ure temperat
umber, D
a
.
2) physical
long the low
point will ap
nd cannot be
he ignition D
at ignition is
chemical re
. The region
Numerical r
H
3
, HO
2
an
at ignition
e distance, L
ture, T
u
= 298
The S-curve
lly unstable
wer weakly
appear at wh
e transport a
Damkohler n
s achieved w
eaction rates
n is defined a
results were
d temperatu
n state com
L = 2.2 cm,
8 K and velo
e consists of
e branch, an
reacting br
hich the chem
away in a st
number, D
a,I
.
when D
a
= D
a
s increase an
as ignition k
e used to inte
ure profile o
mputed usin
unburned n
ocity, V
u
= 10
6
f three
nd 3)
ranch,
mical
teady
The
a,I
.
nd the
kernel
erpret
of
ng
n-
00
te
st
sh
p
lo
fo
st
T
Figure 1
The conc
emperature p
tream at ign
hown in Fig
eak nearly a
ocated at the
or either pre
tagnation pla
The peak reac
1.2 Species
premix
JetSur
2.2 cm
and ve
centration p
profiles are p
nition states
gure 1.1 and
at the same
e middle of t
emixed or n
ane because
ction rate lo
s profiles of H
xed n-dodec
rf 1.0. (Fuel
m, unburned
elocity, V
u
= 1
profiles of
plotted as a
s for both p
Figure 1.2
spatial loca
the domain w
non-premixed
of the expo
ocations are i
H, OH, CH
3
,
cane flames
mole fractio
n-dodecane/
100 cm/s.)
four key s
function of
premixed an
respectively
ation around
which is x =
d flames are
onenetial dep
illustrated in
, HO
2
and te
at ignition
on, X
F
= 10 %
/N
2
mixture
species, H,
the spatial d
nd non-prem
y. It can be
d x = 1.2 cm
= 1.1 cm. Th
e located on
pendence of
n Figure 1.3
emperature p
n state com
%, separate
temperature
OH, CH
3
,
distance from
mixed n-dode
seen that th
m and the sta
herefore, the
n the hot st
reactin rates
and Figure
profile of non
mputed usin
distance, L
e, T
u
= 298 K
and HO
2
,
m the exit of
ecane flame
he active rad
agnation pla
e ignition ke
tream side o
s on tempera
1.4 for prem
7
n-
ng
=
K
and
f fuel
es, as
dicals
ane is
ernels
of the
ature.
mixed
an
lo
ra
w
nd non-prem
ocation of hi
adical pool.
will be discus
Figure 1
mixed n-dod
ighly reactiv
The detaile
ssed in Chap
1.3 Reactio
flames
separa
temper
decane flam
ve region is c
ed analyses
pters 4 and 7
on rates an
at ignition
ate distance,
rature, T
u
= 2
mes respectiv
consistent wi
of the ignit
.
nd temperatu
n state com
L = 2.2 cm
298 K and ve
vely. The r
ith the peak
tion of non-p
ure profile
mputed using
m, unburned
elocity, V
u
=
results dem
concentratio
premixed n-
of premixed
g JetSurf 1
d n-dodecan
100 cm/s.)
monstrate tha
on location o
-dodecane fl
d n-dodecan
1.0. ( = 1.0
e/air mixtur
8
at the
of the
lames
ne
0,
re
1
an
fu
n
m
hy
an
Figure 1
.2 Motiv
Although
nd alternativ
ully understo
eat compon
more approp
ydrocarbons
nd iso-alkan
1.4 Reactio
flames
fractio
dodeca
cm/s.)
vation and O
there are se
ve jet fuels a
ood due to
nents and co
priate surrog
s generally a
ne with carb
on rates and
at ignition
on, X
F
= 10
ane/N
2
mixtu
Objectives
everal studie
and their sur
notably diff
ompositions.
gate compo
above C
4
, i.e
bon number
d temperatur
state comp
%, separate
ure temperatu
es on the com
rrogates, the
ferent oxida
With the
onents are i
e. cyclo-alka
r equivalent
re profile of n
puted using
e distance, L
ure, T
u
= 298
mbustion ch
ey are still no
ative and py
developmen
introduced.
anes with lo
to or large
non-premixe
JetSurf 1.0
L = 2.2 cm,
8 K and velo
haracteristics
ot yet well c
yrolytic char
nt of surrog
Higher m
ong normal-a
er than 8, a
ed n-dodecan
0. (Fuel mol
unburned n
ocity, V
u
= 10
s of convent
characterized
racteristics o
gates of jet f
molecular w
alkyl substit
are considere
9
ne
le
n-
00
tional
d and
of the
fuels,
weight
tuents
ed as
10
essential surrogate components of jet fuels. Some of those neat hydrocarbons may be not
readily available or are very expensive.
In high-speed air-breathing propulsion, heavy hydrocarbons are not used only as fuels
but also as coolants for the fuselage and internal-flow passages. Thus mixtures of light
and heavy hydrocarbons may be involved in the combustion process. The accurately
description of the combustion response of those fuel blends is essential to the engine
design and operation.
Ignition characteristics are important fundamental flame properties towards the
development and validation of chemical kinetics and transport properties and need to be
matched for practical fuels besides other physical and chemical properties, such as the
H/C ratio, other global flame properties, heat release, sooting tendency, flammability, and
regression rate.
In light of the considerations given above, this study aims to achieve systematic
fundamental understanding of the ignition characteristics of both light (gaseous) and
heavy (liquid) hydrocarbons given their importance towards the accurate description of
the combustion behavior of practical fuels. Specifically, the main goal of this
investigation was to provide insights into the ignition characteristics of flames of jet fuels
and related neat hydrocarbons and their mixtures. A high-quality experimental system
was developed achieving well-distributed temperature and flow field assuring numerical
code adapted to experiments. Flame ignition temperatures, T
ign
, were measured in the
counterflow configuration over a range of fuel to N
2
ratio for non-premixed flames and
equivalence ratios, , for premixed flames. All measurements were carried out at
11
atmospheric pressure and elevated fuel-carrying stream temperatures, T
u
. The
experimental data were modeled directly using detailed description of chemical kinetics
and molecular transport, allowing thus for in-depth analysis of the details of the
computed pre-ignition controlling physical and chemical mechanisms.
1.3 Organization of Dissertation
In Chapter 2 the experimental apparatus and procedures are outlined. Temperature
profiles were carefully measured to assure the ability of the system to produce data of
fundamental value. The determination of T
ign
in experiments is introduced as well as the
experimental uncertainties that are discussed in detail.
In Chapter 3, the numerical approach is discussed. The determination of T
ign
through
S-curve marching and the attendant code used to model T
ign
are introduced. The effect of
mixture-averaged and multicomponent transport formulations on flame ignition is
discussed. In addition, the crucial effect of local strain rate, K, on ignition is assessed.
In Chapter 4, the ignition characteristics of premixed flames of gaseous fuel blends
flames are investiogated, with particular interest on the kinetic interactions between fuels.
In Chapter 5, an experimental and numerical study on the ignition of non-premixed
C
3
– C
12
n-alkane flames is presented. Ignition kinetics of n-dodecane flames are
analyzed and compared with that of propane, n-pentane, n-hexane, n-heptane, n-octane,
n-nonane, and n-decane.
12
In Chapter 6, a study is presented on the ignition of non-premixed flames of octane
and decane isomers including 3-methylheptane, 2,5-dimethylhexane, iso-octane, n-octane,
2,7-dimethyloctane, and n-decane.
In Chapter 7, an investigation on the ignition of non-premixed cyclohexane and
mono-alkylated cyclohexane (including methyl-, ethyl-, n-propyl-, and n-butyl-
cyclohexane) flames is presented. The effect of the fuel structure is examined through
detailed numerical simulations. The results of cyclohexane are compared with n-hexane,
and the physical and chemical properties of mono-alkylated cyclohexane are compared
with emphasis on the effects of alkyl substitution (1 to 4 carbons) on the oxidation of
cyclo-alkanes.
In Chapter 8, the ignition characteristics of non-premixed binary ethylene/n-dodecane
fuel blends flames are reported. The role of ethylene addition in facilitating n-dodecane
non-premixed flame ignition and interaction of dominant chemistry and diffusion of
binary fuel blends are studied.
In Chapter 9, a study on the ignition of flames of conventional and alternative jet
fuels is presented. The ignition propensities of these fuels were compared to each other
and notable discrepancies were identified and explained.
In Chapter 10, quasi-empirical correlation formulas of ignition mixing rules for
binary fuel blends were derived for both homogeneous systems and flames.
Finally, the contributions and findings of this dissertation are summarized in Chapter
11. Recommendations for future work are discussed as well.
13
1.4 References
[1] U.S. Energy Information Administration, Annual Energy Outlook 2012.
[2] T.G. DuBois, S. Nieh, Fuel 90 (2011) 1439-1448.
[3] B. Cheze, P. Gastineau, J. Chevallier, Energy Policy 39 (2011) 5147-5158.
[4] L.Q. Maurice, H. Lander, T. Edwards, W.E. Harrison III, Fuel 80 (2001) 747-756.
[5] M.L. Huber, B.L. Smith, L.S. Ott, T.J. Bruno, Energy and Fuels 22 (2) (2008)
1104-1114.
[6] C. Ji, Y.L. Wang, F.N. Egolfopoulos, J Propul Power 27 (2011) 856-863.
[7] L. Rye, C. Wilson, Fuel 96 (2012) 277-283.
[8] C. Guéret, M. Cathonnet, J.C. Boettner, F. Gaillard, F. Gallard, Proc. Combust.
Inst. 23 (1991) 211-216.
[9] A. Violi, S. Yan, E.G. Eddings, A.F. Sarofim, Combust. Sci Technol 174 (2002)
399-417.
[10] C.J. Montgomery, S.M. Cannon, M.A. Mawid, B. Sekar, 40th AIAA Aerospace
Sciences Meeting and Exhibit AIAA 2002-0036, Reno, Nevada ,14-17 January
(2002).
[11] P. Dagaut, Phys. Chem. Chem. Phys. 4 (2002) 2079-2094.
[12] M.A. Mawid, T.W. Park, B. Sekar, C. Arana, 38th Joint Propulsion Conference
and Exhibit AIAA 2002-3876, Indianapolis, Indiana, (2002) 7–10 July.
[13] A. Agosta, N.P. Cernansky, D.L. Miller, T. Faravelli, E. Ranzi, Exp Therm Fluid
Sci 28 (2004) 701-708.
14
[14] J.A. Cooke, M. Bellucci, M.D. Smooke, A. Gomez, A. Violi, T. Favarelli, E.
Ranzi, Proc. Combust. Inst. 30 (2005) 439–446.
[15] E.G. Eddings, S. Yan, W. Ciro, A.F. Sarofim, Combust. Sci Technol 177 (2005)
715-739.
[16] P. Dagaut, M. Cathonnet, Prog. Energy Combust. Sci. 32 (2006) 48-92.
[17] S.S. Vasu, D.F. Davidson, R.K. Hanson, Combust Flame 152 (2008) 125-143.
[18] C.K. Law, Combustion Physics, Chapter 8.
[19] D.F. Davidson, D.R. Haylett, R.K. Hanson, Combust Flame 155 (2008) 108-117.
[20] H. Wang, M.A. Oehlschlaeger, Fuel 98 (2012) 249-258.
[21] D. Darcy, M. Mehl, J.M. Simmie, J. Würmel, W.K. Metcalfe, C.K. Westbrook,
W.J. Pitz, H.J. Curran, Proc. Combust. Inst. 34 (2012)
http://dx.doi.org/10.1016/j.proci.2012.06.131.
[22] K. Kumar, C-J. Sung, Fuel 89 (2010) 2853-2863.
[23] K. Kumar, C-J. Sung, Combust Flame 157 (2010) 676-685.
[24] C. Allen, E. Toulson, T. Edwards, T. Lee, Combust Flame 159 (2012) 2780-2788.
[25] D.B. Lenhert, D.L. Miller, N.P. Cernansky, Combust. Sci Technol 179 (2007)
845-861.
[26] K. Eisazadeh-Far, F. Parsinejad, H. Metghalchi, Fuel 89 (2010) 1041-1049.
[27] J.L. Convery, G.L. Pellett, W.F. O’Brien, L.G. Wilson, 41st Joint Propulsion
Conference and Exhibit AIAA 2005-3776, Tucson, Arizona, 2005 10-13 July.
[28] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, T. Edwards, Proc.
Combust. Inst. 31 (2007) 1205-1213.
15
[29] S. Humer, A. Frassoldati, S. Granata, T. Faravelli, E. Ranzi, R. Seiser, K. Seshadri
Proc. Combust. Inst. 31 (2007) 393-400.
[30] K. Kumar, C.J. Sung, X. Hui, 47th AIAA Aerospace Sciences Meeting AIAA
2009-991, Orlando, Florida, 2009 5–8 January.
[31] T.G. Kreutz, C.K. Law, Combust Flame 104 (1996) 157-175.
[32] T.G. Kreutz, C.K. Law, Combust Flame 114 (1998) 436-456.
16
Chapter 2
Experimental Approach
2.1 Experimental Apparatus
Experiments were carried out under atmospheric pressure in the counterflow
configuration (e.g., [1-6]), as schematically shown in Figure 2.1. The apparatus consists
of an upper quartz burner, from which a hot air jet is directed downward counterflowing
against a pre-vaporized fuel/N
2
jet exiting from the lower burner. In the center of the
quartz burner, a silicon carbide spiral-heating element is embedded and capable of
achieving air temperatures at the burner exit, T
air
, up to 1400 K. The heating of the
element was controlled accurately using a variable AC output transformer (TDGC-3KM).
In order to keep fuel from condensing, the fuel/N
2
jet was heated using a ceramic heating
jacket and maintained at the exit of the lower burner an unburned reactant temperature
constant for all cases considered in each chapter. Both burners included several internal
and external heating capabilities. Additionally, top-hat exit velocity profiles at each
burner exit were achieved by placing honeycombs (400 CPSI) near the exit of the upper
17
burner and by using an aerodynamically contoured nozzle in the lower burner. Burners
with 22 mm diameter nozzles were used, and the nozzle separation distance of 20 mm
and 22 mm were employed.
Figure 2.1 Schematic of the experimental configuration.
Liquid fuels were vaporized using a vaporization system (e.g., [7]) that consists of a
high-precision syringe pump that injects the fuel into a quartz nebulizer. The liquid fuel
was subsequently injected though the nebulizer as a fine aerosol into a crossflow of
heated N
2
. To prevent fuel condensation, the pre-vaporized fuel/N
2
delivery line was
wrapped with heating tape regulated by a temperature controller (OMEGA CN9000A).
The wall temperature of the supply line was monitored by a K-type unsheathed
thermocouple kept approximately 20 K above the boiling point of the fuel to ensure
complete vaporization, and below 473 K to avoid fuel cracking [4]. An R-type
18
unsheathed thermocouple (Pt/13%Rh-Pt) with a bead diameter of about 114 micrometer
was used to measure T
u
and the heated air temperature at the center of exits of the lower
and upper burners respectively. The thermocouple wire was elaborately installed with a
spring at both ends achieving locking force that keeps the wire tight during all the
measurements. A loose wire can cause erratic temperature readings due to the bead
position change. The schematic of the thermocouple is described in Figure 2.2. Before
each experiment, the entire system was allowed to reach steady state for about two hours.
Figure 2.2 Schematic of the thermocouple.
The flow rates of air, N
2
, and gas fuels were metered using sonic nozzles by
controlling the upstream pressure of the orifice. Laser Doppler Velocimetry (LDV) was
used to measure the axial flow velocities along the stagnation streamline. The flow was
seeded with 0.3- m diameter silicon oil droplets produced by a nebulizer [3,5]. The
maximum gradient of the axial velocity profile along the stagnation streamline was
19
defined as a characteristic local strain rate K, i.e. K ( u / z )
max
where u is the axial
velocity and z is the axial distance from the lower burner exit. In the present experiments,
K had to be measured on the fuel-side, as the silicon droplets cannot survive the high
temperatures encountered in the upper quartz burner, and in modeling the flames that was
taken into account. The local strain rate was measured and it was determined to be nearly
constant in groups of experiments in the following chapters for the comparison of
ignition characteristics of the certain class of fuels.
2.2 Experimental Procedures
Ignition was achieved by establishing the appropriate flow rates and temperatures at
the two burner exits, and by gradually increasing the fuel flow rate until ignition was
observed. This approach was preferred over raising the air temperature until ignition
happens, as the uncertainty in the fuel flow rates is notably lower compared to that of the
air temperature. Thus, for a given air temperature, the ignition propensity of the various
fuels was quantified with great relative accuracy limited by that of the flow rate.
Considering that the thermocouple could disturb the flow and result in unwanted local
mixing and thus in premature ignition, it was retracted from the test section before
injecting the fuel after the desired temperature of the upper burner exit was established.
2.2.1 Determination of Ignition Limits
The radial temperature profiles were measured at different distances from the upper
burner exit and the results are shown in Figure 2.3. Results indicate that the temperature
20
profiles are rather uniform for radii up to 7 mm around the centerline of the stream,
which assures that the assumptions invoke in the numerical simulations are satisfied.
Obtaining radial temperature uniformity for the most part of the test section is
challenging, and it was achieved after extensive modifications and improvements of the
heating and insulating configurations of the quartz burner. The measured T
air
at the
center of the burner exit at the state of ignition was defined (e.g., [6,8,9]) as the ignition
temperature, T
ign
, once corrected for radiation/convective heat transfer [8]. The range of
T
ign
considered in this study was mainly 1223 – 1443 K.
Figure 2.3 Radial temperature profiles measured at different distances, z, from
the upper burner exit; ( ◆) z = 0 mm; ( ■) z = 2 mm; ( ◇) z = 4 mm;
( ▲) z = 6 mm; ( □) z = 8 mm.
900
950
1000
1050
1100
1150
1200
1250
1300
1350
1400
-11 -9 -7-5-3 -1 1 3 5 7 911
Radius, mm
Temperature,K
centerline
21
T
ign
’s monotonically increase with increasing K [8]. At low strain rate, T
ign
is low,
whereas T
ign
is much higher at high strain rate. In the strain rate range (from 120 s
-1
to
200 s
-1
) considered in the present study, T
ign
’s are mostly above 1200 K, for which
ignition is controlled by high temperature kinetics. Simulation results calculated for the
typical fuels in different chemical class, i.e. n-dodecane vs. n-pentane and n-hexane vs.
cyclohexane, show that under a wide range of strain rates (from 100 s
-1
to 800 s
-1
) the
hierarchy of ignition temperatures does not change. For low strain rates, since T
ign
is low,
there could be new oxidation pathways of fuel, which involve the low to intermediate
temperature chemistry [10]. In order to validate the ignition behavior at ultra-low strain
rates, it requires additional experimental data and reliable detailed kinetic models for low
and intermediate temperatures. This is recommended for the future work. The hierarchy
of ignition temperatures of the fuels depends on the chemical kinetics that control
ignition at their corresponding temperature range from very low to very high strain rates.
2.2.2 Vaporization of Liquid Fuels
The vaporization of liquid fuels and their complete mixing with gaseous inert or
oxidizing streams are two important steps in the experiments where the characteristic
time and geometric space are limited. The vaporization system produces small droplets
of the liquid fuel and introduces them into the gaseous medium to achieve full
vaporization and mixing. Local high concentration of the fuel resulting from incomplete
vaporization or inadequate mixing will lead to non-reliable ignition conditions. Under a
22
specific T
u
, the Antoine equation was employed for the estimation of vapor pressure, P
v
,
and given below.
(2.1)
Where A, B, and C are experimentally determined component-specific constants and T is
the prevailing temperature. The partial pressure of the liquid fuel was kept below P
v
to
avoid the condensation in the line downstream of the vaporization system at the
prevailing temperature and pressure.
The low vapor pressure of heavy liquid fuels limits the range of fuel concentrations
that could be examined experimentally without experiencing fuel condensation in cold
spots and/or thermal cracking in hot spots. In the present study, the fuel mole fraction
range in the fuel/N
2
mixture, X
F
, was varied from 1% to 12%, and expressed as:
where N is the number of moles, and subscript Fuel and N
2
represents fuel and nitrogen
in the fuel/N
2
mixture, respectively.
2.3 Experimental Uncertainties
In the present experiments, a thermocouple was used to measure the high
temperatures of the air stream that were as high as 1400 K. Radiation and convection
heat transfer occurs largely around the bead of thermocouple, eventually reaching a
thermal balance, which constitute the main uncertainty of the determination of ignition
23
temperatures. The reading temperature of the thermocouple has to be corrected
accordingly.
The calculation procedures presented by Fotache et al. [8] was adopted herein for
making the temperature corrections. The value of corrected ignition temperature depends
on the choice of flow models, specifically the sphere and cylinder in cross-flow,
describing the flow around the bead of thermocouple. By considering the geometry of
the thermocouple, the average value of the corrected temperatures using two flow models
around the thermocouple bead was chosen to be best fit with the experimental
measurement, resulting in air temperature uncertainty that varied between 11 and 16
K; the larger uncertainty corresponds to the higher temperatures. Two thermocouples
were employed for comparison of the same experimental measurement to avoid
temperature errors resulting from the wear of the material of the thermocouple that could
slowly deteriorate due to oxidation or even be broken by unexpected saturating in the
flames.
For the other uncertainty sources, the effect of the value of the emissivity of
thermocouple bead,
b
, on the determination of the corrected temperature cannot be
neglected comparing to the emissivities of quartz burner and heating element, leading to
an uncertainty of 5 K.
Based on the thermocouple readings, the fluctuation of the air temperature at the exit
of the upper burner was estimated to be within 3 K when the operation of the silicon
carbide spiral-heating element was stable. The above involved uncertainties were
responsible for the absolute uncertainty of ignition temperature, which is very important
24
when comparing the experimental data with the computed results, or in other words,
validating kinetic models.
The relative uncertainties were caused by the repeatability of the measurement of
ignition temperatures at the same data point, and were determined within 5 K, which
could be aptly minimized by careful experimental procedures. A temperature rise of 20
K at hot boundary was set to allow for the definitive comparision of adjacent data points.
The small size of relative uncertainty bars allows for specifying the hierarchy of ignition
temperatures of different fuels.
The T
u
values measured at the center of the lower burner exit, from which the fuel/N
2
mixture was supplied, fluctuated within 2 K.
The upstream pressure of each sonic nozzle was monitored by a pressure gauge with
an uncertainty of ±0.25%. The syringe pump used to inject the liquid fuels has discrete
setting values. The injection rates of the liquid fuels were chosen to be very close to the
ignition state until the uncertainty of the mole fraction of fuel in the fuel/N
2
mixture at the
ignition state was within % 2 . The LDV velocity measurements are accurate within
±1%. The 2 uncertainty of K was determined to be ±10 s
-1
, where is the standard
derivation and expressed as
N
i
i
N K K
1
2
) 1 /( ) (
.
In the expression, K is the mean stain rate, K
i
is the ith measurement of stain rate, and N
is the total number of data sets.
25
2.4 References
[1] C.K. Wu, C.K. Law, Proc. Comb. Inst. 20 (1985) 1941-1949.
[2] G. Yu, C.K. Law, C.K. Wu, Comb. Flame 63 (1986) 339-347.
[3] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, Combust. Flame 144
(2006) 448-460.
[4] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, T. Edwards, Proc.
Combust. Inst. 31 (2007) 1205-1213.
[5] C. Ji, E. Dames, Y.L. Wang, H. Wang, F.N. Egolfopoulos, Combust. Flame 157
(2010) 277–287.
[6] N. Liu, C. Ji, F.N. Egolfopoulos, Combust. Flame 159 (2012) 465-475.
[7] Y.L. Wang, Q. Feng, F.N. Egolfopoulos, T.T. Tsotsis, Combust. Flame 158 (2011)
1507-1519.
[8] C.G. Fotache, T.G. Kreutz, D.L. Zhu, C. K. Law, Combust. Sci. Technol. 109
(1995) 373–393.
[9] C.G. Fotache, H. Wang, C. K. Law, Combust. Flame 117 (1999) 777-794.
[10] R. Grana, K. Seshadri, A. Cuoci, U. Niemann, T. Faravelli, E. Ranzi, Combust.
Flame 159 (2012) 130-141.
26
Chapter 3
Numerical Approach
3.1 Codes Description
Simulations were performed using an opposed-jet flow code [1], in which quasi-1-D
steady conservation equations of mass, momentum, species concentrations, and energy
along the stagnation streamline are solved. Kee and coworkers [2] have developed the
original version of the code. Additionally, the code has been modified to allow for any
type of boundary condition, steady or unsteady, including the effect of thermal radiation
from CH
4
, CO, CO
2
, and H
2
O at the optically thin limit [1], and it is integrated with the
CHEMKIN [3] and Sandia Transport [4] subroutine libraries. The H and H
2
diffusion
coefficients of several key pairs are based on the recently updated set of Lennard-Jones
(L-J) parameters [5,6].
The ignition S-curve [7] was reproduced by monitoring the variation of the maximum
concentration of any radical with the temperature of the hot boundary, referred herein as
the heated stream at the upper burner exit, as shown in Figure 3.1. T
ign
is determined as
27
the hot boundary temperature, T
hot boundary
, at the lower turning point of the S-curve, which
corresponds to the transition to a vigorously burning flame. To capture the turning-point
behavior around ignition and extinction states, the stagnation-flow code was modified by
introducing the methodology of internal boundary [8,9]. Specifically, one-point
continuation [10] was implemented that a pre-determined species concentration imposed
at the location where the species concentration exhibits a maximum slope serves as the
new boundary condition replacing T
hot boundary
to
obtain single-valued behavior.
Figure 3.1 Schematic of a typical S-curve determined in the opposed-jet
configuration.
The experimental and numerical determined T
ign
’s are compatible, as a result of the
same definition as the T
hot boundary
’s. And such definition is physically reasonable. In the
reaction zone, diffusion balances reaction. Given that the location of maximum diffusion
is
al
to
T
m
co
n
s at the loca
lso the locat
o the hot bou
T
T
The mole
mixture-avera
ode allows
otably highe
Figure 3
tion at whic
ion where ig
undary temp
.
cular transp
aged [4] and
for. The
er compared
3.2 Compu
bound
multico
X
F
= 5.
ch the secon
gnition occur
erature and
ort propertie
d multicomp
computation
to using the
uted respons
ary tempe
omponent fo
.10 %, T
u
= 4
nd derivative
rs. The actu
T
T
es of gaseou
onent [11] tr
nal time usi
e mixture-ave
ses of maxim
erature by
ormulations
401 K, K = 13
e of tempera
ual T
ign
in the
/ T
us mixture w
ransport coe
ing the mul
eraged form
mum H radica
y using
for a non-pr
30 s
-1
.
ature is the m
e ignition ke
T
were compute
efficient form
lticomponen
mulation.
al mass fract
mixture-ave
remixed iso-
maximum, th
ernel is very
≪1. Ther
ed using bot
mulations tha
nt formulatio
tion to hot ai
eraged an
-octane flame
28
his is
close
refore,
th the
at the
on is
ir
nd
e,
29
The differences between T
ign
’s results for n-dodecane flames obtained using the two
formulations are, for example, as high as 16 K [12]. Figure 3.2 shows a 13 K difference
for an iso-octane flame at X
F
= 5.10 %, T
u
= 401 K and K = 130 s
-1
. Therefore, it is
worthwhile using the multicomponent formulation when considering high accuracy. All
computed T
ign
’s that are reported in this study, were determined using the multi-
component transport coefficient formulation.
The number of grid points in the simulations was varied between 300 and 800 for the
various conditions considered. Grid refinement was terminated when the effect on T
ign
was less than 1 K.
The effects of chemical kinetics and molecular diffusion on T
ign
were assessed
through sensitivity analysis. Dong et al. [5] and Holley et al. [13] have modified the
opposed-jet code in order to allow for the mathematically rigorous sensitivity analyses of
all dependent variables, including T
ign
, to all reaction rate constants as well as to all
binary diffusion coefficients (BDC). This is possible because in the continuation
approach invoked in the code, T
ign
becomes a dependent variable and thus rigorous
sensitivity analysis can be performed on them. More specifically, the logarithmic
sensitivity coefficients (LSC) of T
ign
on reaction rate coefficients and BDC’s are
formulated as
i ign
A d T d LSC ln / ln and
m k ign
D d T d LSC
,
ln / ln respectively, in
which A
i
is the pre-exponential factor of Arrhenius equation of reaction i, and D
k,m
is the
BDC between species k and m. It should be noted that a negative LSC value implies a
positive effect on ignition, as T
ign
tends to decrease.
3
th
pr
ca
v
sh
ex
si
o
o
th
.2 Effect
In order f
he local stra
resence of th
an introduce
elocities alo
hould be not
xtent of hea
ignificant ef
f the velocit
f the domai
heir values c
Figure 3
ts of Local S
for the simu
ain rates ado
hermal and m
e small but
ong the strea
ted that the
at and radica
ffect on the d
ties at the tw
n, L, and th
an influence
3.3 Axial v
side.
Strain Rate
lations to re
opted in the
momentum b
finite two-
amline does
local strain
al losses fro
determinatio
wo burner ex
he velocity g
e the distribu
velocity pro
epresent the
simulations
boundary lay
-dimensional
not warran
rate just ups
m the igniti
on of therma
xits and two
gradient at b
ution of the s
file along th
experiments
and the exp
ayers in coun
l effects an
nt similar va
stream of the
ion kernel (e
al ignition te
geometric p
burner exit,
strain rate w
he stagnation
s properly, i
periments ar
nterflow igni
nd as a resu
alues of loca
e ignition ke
e.g., [14-20]
emperatures.
parameters,
are notab
ithin the ign
n streamline
it is essentia
re matched.
ition experim
ult, matching
al strain rate
ernel control
]) and thus
. The magn
separate dis
bly importan
nition kernel
e on the fue
30
al that
The
ments
g the
es. It
ls the
has a
nitude
stance
nt, as
[1].
l-
31
In the simulations, the stretch-corrected T
ign
’s were obtained by changing the value of
the velocity gradient at the burner exit to conform it with the experimental strain rate
distribution, as shown in Figure 3.3. More specifically, for a non-premixed iso-octane
flame at X
F
= 5.10 %, T
u
= 401 K, T
ign
’s were computed at = 0 s
-1
and = 60 s
-1
for the
same experimental velocities at two burner exits. Note that strain rate corrections
resulted in the reduction of the computed T
ign
by as much as 23 K, indicating a rather
strong dependence of T
ign
on boundary conditions. Therefore, for all simulations, the
experimental conditions, especially velocities at two burner exits, L and , were carefully
recorded and adopted.
3.3 Kinetic Models
In this work a variety of detailed chemical kinetics models were implemented in the
modeling of the data along with the attendant thermodynamic and transport properties.
The most recent models have been chosen, which have been validated against data
obtained in homogeneous reactors and flames.
The influence of kinetics and molecular transport was assessed through sensitivity
and reaction path analyses utilizing the computed flame structures.
3.4 References
[1] F.N. Egolfopoulos, Proc. Comb. Inst. 25 (1994) 1375-1381.
[2] R.J. Kee, J.A. Miller, G.H. Evans, G. Dixon-Lewis, Proc. Comb. Inst. 22 (1989)
1479-1494.
32
[3] R.J. Kee, F.M. Rupley, J.A. Miller, Chemkin-II: A Fortran Chemical Kinetics
Package for the Analysis of Gas-Phase Chemical Kinetics, Sandia Report
SAND89-8009, Sandia National Laboratories, 1989.
[4] R.J. Kee, J. Warnatz, J.A. Miller, A FORTRAN Computer Code Package for the
Evaluation of Gas-Phase Viscosities, Conductivities, and Diffusion Coefficients,
Sandia Report SAND83-8209, Sandia National Laboratories, 1983.
[5] Y. Dong, A.T. Holley, M.G. Andac, F.N. Egolfopoulos, S.G. Davis, P. Middha, H.
Wang, Combust. Flame 142 (2005) 374–387.
[6] P. Middha, H. Wang, Combust. Theor. Model. 9 (2005) 353-363.
[7] C.K. Law, Combustion Physics, Cambridge University Press, New York, 2006
(Chapter 8).
[8] M. Nishioka, C.K. Law, T. Takeno, Combust. Flame 104 (1996) 328-342.
[9] F.N. Egolfopoulos, P.E. Dimotakis, Proc. Comb. Inst. 27 (1998) 641-648.
[10] M.G. Andac, F.N. Egolfopoulos, Proc. Comb. Inst. 31 (2007) 1165-1172.
[11] G. Dixon-Lewis, Proc. Roy. Soc. A 307 (1968) 111-135.
[12] N. Liu, C. Ji, F.N. Egolfopoulos, Combust. Flame 159 (2012) 465-475.
[13] A.T. Holley, X. You, E. Dames, H. Wang, F.N. Egolfopoulos, Proc. Combust. Inst.
32 (2009) 1157-1163.
[14] T.G. Kreutz, M. Nishioka, C.K. Law, Combust. Flame 99 (1994) 758-766.
[15] C.G. Fotache, T.G. Kreutz, D.L. Zhu, C. K. Law, Combust. Sci. Technol. 109
(1995) 373–393.
[16] C.G. Fotache, T.G. Kreutz, C. K. Law, Combust. Flame 110 (1997) 429-440.
33
[17] C.G.Fotache, T.G. Kreutz, C. K. Law, Combust. Flame 108 (1997) 442-470.
[18] T.G. Kreutz, C.K. Law, Combust. Flame 114 (1998) 436-456.
[19] C.G. Fotache, H. Wang, C. K. Law, Combust. Flame 117 (1999) 777-794.
[20] C.G. Fotache, Y. Tan, C.J. Sung, C. K. Law, Combust. Flame 120 (2000) 417-426.
34
Chapter 4
Ignition Characteristics of Premixed Alternative Gaseous Fuel Blends
Flames
4.1 Introduction
Energy conversion and power generation are notably affected worldwide by a series
of problems ranging from fuel prices, to environmental regulations, and to world politics.
Thus, alternative sources of energy are attracting notable interest due to their petroleum
and natural gas independent, as well as environmental benefits.
Alternative gaseous fuels, produced from the utilization of coal, landfills, and
biomass, perform a highly efficiency and low emission comparing to the direct
combustion of themselves in power generation, noted not only by commercial and
government sectors. Such fuels including gasified biomass, gasified coal gas, and syngas
are renewable and with short carbon cycle. At the same time there is also interest in
utilizing blended fuels, with sources ranging from natural gas, process and refinery gases,
to landfill gases in IGCC and other applications, including industrial boilers and gas
reciprocating engines.
35
The flame properties of a particular fuel can be sensitive strongly to composition
variations as well as the presence of C
1
-C
4
hydrocarbons [1]. Syngas contains largely
two particularly clean-burning fuels, CO and H
2
, with variable compositions depending
on the coal types and the post-processing technique. Landfill gas may contain notable
amounts of CO
2
and gaseous hydrocarbons like methane, ethane, and propane, depending
on the type of waste used [2]. Associated gases are common in oil field operations, of
which global combustion characteristics may be similar to those of natural gases but
emissions and other characteristics such as lean limits and combustion instabilities might
be significantly different. Coke oven gas containing high H
2
content is a medium heating
value gas fuel, of which composition varies due to the coal ingredients and coking
temperature. The investigations on the flame properties of such types of fuel blends are
of important in combustor design and emission controls.
Detailed kinetic models have been proposed for the oxidation of H
2
, CO, and C
1
-C
4
hydrocarbons (e.g., [3-9]) and validated generally based on flow reactor and laminar
flame speed experiments. Extensively experimental studies of combustion characteristics
have been done for neat component fuels of H
2
, CO, and C
1
-C
4
hydrocarbons (e.g., [10-
12]) and several investigations are for the mixtures (e.g., [13]). It is worthwhile to note
that ignition characteristics studies by appropriately combination of kinetics and transport
effects are largely been done for non-premixed flames (e.g., [14-17]) while those for
premixed flames are scant [17-19].
36
The goal of the present study was to investigate the premixed flame ignition
characteristics of some of the aforementioned fuel blends due to its importance to gas
turbines applications.
4.2 Experimental Approach
The experiments were performed in the counterflow configuration, as shown in
Figure 4.1. A hot N
2
stream was directed downward from the upper quartz burner against
the fuel/air stream exiting from the lower burner. Both the burner nozzle diameters and
the burner separation distance were 22 mm. The neat fuel components and air were
mixed in a mixing chamber before injecting into the lower burner and their flow rates
were accurately controlled by the mass flow meters. Under all experimental conditions,
T
u
= 423 K and K = 200 s.
-1
Figure 4.1 Schematic of the experimental configuration.
37
Ignition was achieved by establishing the appropriate boundary temperatures at the
center of two burner exits and by gradually increasing the flow rate of H
2
when keeping
the flow rates of the other components constant until a flame appeared. The N
2
temperature at the center of top burner exit, T
N2
, was defined as T
ign
herein after the
correction for radiation/convective heat losses.
Measurements were performed for mixtures of air with the fuel blends shown in
Table 4.1.
Table 4.1 Composition of Blended Fuels.
Fuel
No.
Mole %
Classification or Application
CH
4
C
3
H
8
CO
2
H
2
CO
1 30 0 0 60 10 Coke Oven Gas
2 5 0 0 95 0 Chemical Process Byproduct Gas
3a 10 0 0 32 58 Syngas
3b 0 10 0 32 58
4 25 0 10 54 11 Syngas
4.3 Modeling Approach
USC Mech II consisting of 111 species and 784 reactions, developed by Wang et al.
[8], was used for all the simulations in the present Chapter. It describes the high-
temperature kinetics of hydrogen, carbon monoxide, and C
1
-C
4
hydrocarbons, and
incorporates the recent thermodynamic, kinetic, and species transport updates. The
reaction model has been validated against extensive H
2
/CO/C
1
-C
4
combustion data.
4
F
o
F
w
H
.4 Resul
Experime
uel # 3a/air,
for fuel lea
f the data fo
uel # 2 and
was determin
H
2
[20] in all
Figure 4
lts and Disc
ental and com
, and Fuel #
an condition
or all fuel b
Fuel # 3a a
ned also to in
fuel blend c
4.2 Experi
# 1, 3
( ■)(‐ ‐
25% C
ussion
mputed T
ign
4/air are plo
ns. The com
blends. They
and underpre
ncrease only
considered.
imental and
30%CH
4
/60%
‐) Fuel # 3a
CH
4
/10%CO
2
’s for premi
otted in Figu
mputed T
ign
’s
y are in clos
edict slightly
y slightly wit
computed T
%H
2
/10%CO
a, 10% CH
4
/3
2
/54%H
2
/11%
ixed flames
ure 4.2 as a f
using USC
se agreemen
y the data of
th the due
T
ign
’s of prem
O; ( ♦)(∙∙∙) Fu
/32%H
2
/58%
%CO.
of Fuel # 1
function of e
Mech II rep
nt with expe
f Fuel # 1 an
e to the high
mixed flames;
uel # 2, 5%
%CO; ( ▲)(–
1/air, Fuel #
equivalence
produce the
rimental dat
nd Fuel # 4.
h concentrati
; ( ◇)(―) Fue
%CH
4
/95%H
2
∙ ‒) Fuel # 4
38
#2/air,
ratio,
trend
ta for
. T
ign
ion of
el
2
;
4,
39
Logarithmic sensitivity coefficients of T
ign
’s on reaction rate coefficients were
computed for Fuel # 1 at = 0.40 and = 0.61 and shown in Figure 4.3. It can be seen
T
ign
’s are primarily sensitive to chain branching and termination reactions R1 and R2.
H + O
2
→ O + OH (R1)
H + O
2
+ M → HO
2
+ M (R2)
Small effects from CH
4
related reactions on ignition were identified also in Figure 4.3.
CH
4
consumes O/OH/H and produces CH
3
inhibiting thus ignition. According to the
reaction path analysis, the H radical consumption increases slightly as the amount of CH
4
in the fuel/air mixture increases. Thus T
ign
at = 0.61 is slightly higher than that at
= 0.40.
Figure 4.3 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for Fuel # 1 at = 0.40 and = 0.61.
Flames of Fuel # 4 were found to have the highest ignition limits, whereas flames of
Fuel # 2 exhibit the lowest one as shown in Figure 4.2. This is reasonable because Fuel #
-0.14 -0.12 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08
H + O
2
→ O + OH
OH + H
2
→ H + H
2
O
O + H
2
→ H + OH
CH
4
+ H → CH
3
+ H
2
CH
4
+ O → CH
3
+ OH
CH
4
+ OH → CH
3
+ H
2
O
H + O
2
+ M → HO
2
+ M
Fuel # 1
(30%CH
4
/60%H
2
/10%CO)
40
2 has the highest concentration of H
2
that tends to decrease T
ign
’s. Fuel # 1 and Fuel # 3a
consist of the same neat component fuels but different compositions. Comparing those
two fuel blends, T
ign
’s of flames of Fuel #1 are higher than those of Fuel # 3a. The spatial
variations of the fuel concentration at = 0.50 for flames of Fuel # 1 and Fuel # 3a are
shown in Figure 4.4. The concentration of CH
4
in Fuel # 1 is much higher than that in
Fuel # 3a while the H
2
concentration of Fuel # 1 is only slightly higher than that in Fuel #
3a. Such higher percentage of CH
4
in Fuel #1 leads to a higher ignition limits. The
higher T
ign
’s of flames of Fuel #1 compared to Fuel # 3a can be also implicitly elucidated
by the higher concentration of CH
3
and lower amount of H radicals at ignition kernel
shown in Figure 4.5. Fuel # 4 has a similar composition as Fuel # 1 but is diluted by 10%
CO
2
and as a result Fuel # 4 ignites harder than Fuel # 1.
Figure 4.4 Concentration profiles of CH
4
and H
2
for Fuel # 1 and Fuel # 3a at
= 0.50 at ignition state.
Distance from the lower burner, x, cm
Fuel # 1 CH
4
Fuel # 3a CH
4
Fuel #1 H
2
Fuel #3a H
2
Mole Fraction
41
Figure 4.5 Concentration profiles of CH
3
and H for Fuel # 1 and Fuel # 3a at
= 0.50 at ignition state.
Figure 4.6 Experimental and computed T
ign
’s of premixed flames; ( ■)(―) Fuel
# 3a, 10 % CH
4
/ 32 % H
2
/ 58 % CO; ( □)(‐‐‐) Fuel # 3b, 10 % C
3
H
8
/
32 % H
2
/ 58 % CO.
950
970
990
1010
1030
1050
1070
1090
1110
1130
1150
0.45 0.55 0.65 0.75 0.85 0.95
Ignition Temperature, T
ign
, K
Equivalence Ratio,
Fuel # 1 CH
3
Fuel # 3a CH
3
Fuel #1 H*10
Fuel #3a H*10
Mole Fraction
Distance from the lower burner, x,
42
Figure 4.6 depicts the experimental and computed T
ign
’s of flames of Fuel # 3a/air and
Fuel # 3b/air mixtures under fuel lean condition. The simulation results are in close
agreement with the experimental data within the experimental uncertainty, which is
20 K. It can be seen that T
ign
’s of flames of Fuel # 3b are higher than that of Fuel # 3a
by 30 – 35 K. This result is interesting because it is contrary to the expected. According
to the simulations, CH
4
/air flames exhibit an approximately 200 K increase in T
ign
’s
comparing to C
3
H
8
/air flames. However, the hierarchy of T
ign
’s is opposite when mixing
them with H
2
and CO, which inferred the interaction of blended components or their
main intermediates and such was demonstrated by the reaction path analyses.
Figure 4.7 (a) and (b) illustrate the reaction path analyses of flames of C
3
H
8
/air and
Fuel # 3b/air at the ignition state for = 0.7, respectively. Note that H
2
addition in the
fuel blends altered the consumption route of C
2
H
4
. At the ignition state of Fuel # 3b/ air
flames, C
2
H
4
is one key intermediate during the oxidation of C
3
H
8
and subsequently form
CH
3
via R3 and R4 instead of forming C
2
H
3
which is a main product at the ignition of
C
3
H
8
/air flames.
C
2
H
4
+ H + M → C
2
H
5
+ M (R3)
C
2
H
5
+ H → 2CH
3
(R4)
Moreover, the structure of ignition kernel was analyzed and revealed that CH
3
radicals produced from C
3
H
8
is higher comparing to that from CH
4
. CH
3
radicals can
readily recombine H radicals reducing thus the rate of R1. Therefore, C
3
H
8
addition in
the fuel blend increases T
ign
more than the addition of CH
4
. Such results are consistant
with the effects on flame propagation and extinction by Park et al. [21].
43
(a)
(b)
Figure 4.7 Reaction path analyses of flames of (a) C
3
H
8
/air, (b) Fuel # 3b (10 %
C
3
H
8
/ 32 % H
2
/ 58 % CO)/air at the ignition state and = 0.70
using USC Mech II. M represents a third body. The numbers
indicate the conversion percentages.
44
4.5 Concluding Remarks
Ignition temperatures of premixed flames of Fuel # 1 (30%CH
4
/60%H
2
/10%CO) /air,
Fuel # 2 (5%CH
4
/95%H
2
) /air, Fuel # 3a (10%CH
4
/32%H
2
/58%CO) /air, Fuel # 3b
(10%C
3
H
8
/32%H
2
/58%CO) /air, Fuel # 4 (25%CH
4
/10%CO
2
/54%H
2
/11%CO) /air
mixtures were measured in the counterflow configuration under atmospheric pressure, a
constant strain rate of 200 s,
-1
and an unburned fuel/air mixture temperature of 423 K.
Laser Doppler Velocimetry was used to determine the local strain rate on the fuel side.
USC Mech II closely predicts T
ign
’s of flames of Fuel # 2, Fuel # 3a, and Fuel # 3b, and
slightly underpredicts T
ign
’s of flames of Fuel #1 and Fuel # 4.
T
ign
was found to be insensitive to the equivalence ratio due to the high concentration
of H
2
in all fuel blends. The presence of CH
4
in the fuel blends inhibits ignition because
it consumes H/O/OH radicals and produces the relatively stable CH
3
radicals. Flames of
Fuel # 1 exhibit lower ignition propensity comparing to Fuel # 3a because of the higher
concentration of CH
4
. Fuel #3b ignites harder than Fuel # 3a due to the higher
production of methyl radicals from C
3
H
8
comparing to that from CH
4
.
Sensitivity analyses on kinetics revealed that the ignition temperature is primarily
sensitive to the chain branching and termination reactions: (1) H + O
2
→ O + OH and (2)
H + O
2
+ M → HO
2
+ M.
4.6 References
[1] J. Natarajan, S. Nandula, T. Lieuwen, J. Seitzman, Proceedings, ASME Turbo
Expo (2005) 677-686.
45
[2] W. Qin, F.N. Egolfopoulos, T.T. Tsotsis, Chem. Eng. J. 82 (2001) 157-172.
[3] N.M. Marinov, W.J. Pitz, C.K. Westbrook, A.M. Vincitore, M.J. Castaldi, S.M.
Senkan, Combust. Flame 114 (1998) 192-213.
[4] M.A. Mueller, R.A. Yetter, F.L. Dryer, Int. J. Chem. Kinet. 31 (1999) 705-724.
[5] M. Ó Conaire, H. J. Curran, J. M. Simmie, W. J. Pitz, C. K. Westbrook, Intl. J.
Chem. Kinet. 36 (11) (2004) 603–622.
[6] S. G. Davis, A. V. Joshi, H. Wang, F. Egolfopoulos, Proc. Combust. Inst. 30
(2005) 1283-1292.
[7] P. Saxena, P., F.A. Williams, Combust. Flame 145 (2006) 316-323.
[8] Hai Wang, Xiaoqing You, Ameya V. Joshi, Scott G. Davis, Alexander Laskin,
Fokion Egolfopoulos & Chung K. Law, USC Mech Version II. High-Temperature
Combustion Reaction Model of H2/CO/C
1
-C
4
Compounds.
http://ignis.usc.edu/USC_Mech_II.htm, May 2007.
[9] D. Healy, D.M. Kalitan, C.J. Aul, E.L. Petersen, G. Bourque, H. J. Curran, Energy
and Fuels 24(3) (2010) 1521-1528.
[10] F.N. Egolfopoulos, P. Cho, C.K. Law, Combust. Flame 76 (1989) 375–391.
[11] Y. Dong, A.T. Holley, M.G. Andac, et al., Combust. Flame 142 (2005) 374–387.
[12] G. Jomaas, X.L. Zheng, D.L. Zhu, C.K. Law, Proc. Combust. Inst. 30 (2005) 193–
200.
[13] J. Herzler, C. Naumann, Combust. Sci. Tech. 180 (2008) 2015–2028.
[14] C.G. Fotache, T.G. Kreutz, D.L. Zhu, C.K. Law, Combust. Sci. Tech. 109 (1995)
373–393.
46
[15] C.G. Fotache, Y. Tan, C.J. Sung, C.K. Law, Combust. Flame 120 (2000) 417–426.
[16] C. Safta, C.K. Madnia, Combust. Flame 144 (2006) 64–73.
[17] S. Humer, R. Seiser, K. Seshadri, Proc. Combust. Inst. 29 (2002) 1597–1604.
[18] D.G. Vlachos, L.D. Schmidt, R. Aris, Combust. Flame 95 (1993) 313–335.
[19] X.L. Zheng, J.D. Blouch, D.L. Zhu, T.G. Kreutz, C.K. Law, Proc. Combust. Inst.
29 (2002) 1637–1643.
[20] C.G. Fotache, T.G. Kreutz, C.K. Law, Combust. Flame 110 (1997) 429–440.
[21] O. Park, P.S. Veloo, N. Liu, F.N. Egolfopoulos, Proc. Combust. Inst. 33 (2011)
887-894.
47
Chapter 5
Ignition of Non-Premixed C
3
-C
12
n-alkane Flames
5.1 Introduction
Practical fuels, such as gasoline, diesel, and kerosene, are composed of hundreds and
often thousands of hydrocarbons, with n-alkanes being among the dominant components.
Due to the chemical complexity of practical fuels their combustion properties cannot be
modeled from first principle and understood.
Thus, by matching the physical and chemical properties of the practical fuels
surrogate fuels containing a limited number of representative hydrocarbon compounds
can be developed. Due to the relatively small number of neat species present, surrogate
fuels could be modeled in detail. The development and validation of reliable kinetic
models that describe the pyrolysis and oxidation of the candidate neat components and
their mixtures relies critically on the existence of experimental data. Such data include
species profiles and ignition delay times obtained in homogeneous reactors as well as
laminar flame speeds, flame ignition and extinction limits, and flame structures. Among
48
those properties, flame ignition of large molecular weight hydrocarbons in well-
controlled configurations has been studied the least. On the other hand, ignition in well-
mixed spatially homogeneous reactors has been studied extensively and ignition delay
data have been valuable towards the development of detailed kinetic models.
Several studies have been conducted on the autoignition of n-alkanes in shock tubes
and rapid compression machines (RCM).
Burcat et al. [1] studied C
1
-C
5
n-alkanes in a shock tube. It was shown that the
ignition delay times of methane are approximately one order of magnitude greater than
C
2
-C
5
alkanes and that the ignition delay time decreases with increasing number of
carbon atoms in the molecule with the exception of ethane. Experimental studies on
ignition delay times for n-butane have been carried out in RCM’s (e.g. [2-11]) and shock
tubes (e.g. [1,10,12]). Griffiths et al. [8] determined the autoignition characteristics of n-
butane, n-pentane, n-hexane, and n-heptane and it was found that the fuel reactivity
increases with the carbon number for n-alkanes. The autoignition behavior of n-heptane,
being one of the primary reference fuels, has received notable attention during the last
two to three decades (e.g. [12-21]).
Due to limitations stemming from the low vapor pressure, ignition studies for n-
alkanes larger in molecular size than n-heptane are limited. Olchanski and Burcat [22]
investigated ignition delays of mixtures of n-decane/oxygen/argon in a shock tube.
Zhukov et al. [23] reported the ignition delay times of stoichiometric and lean n-decane/
mixtures air, again obtained in a shock tube. Kumar et al. [24] studied the autoignition of
various n-decane/oxidizer mixtures in a RCM. Vasu et al. [25] measured ignition delay
49
times and OH concentration time-histories, and subsequently Davidson et al. [26]
measured concentration time-histories of n-dodecane, C
2
H
4
, OH, CO
2
, and H
2
O during n-
dodecane oxidation behind reflected shock waves. Most recently, Haylett et al. [27]
reported time-history measurements for OH and C
2
H
4
during the oxidation of n-
hexadecane behind reflected shock waves in a new second-generation aerosol shock tube
over a temperature range of 1120 K to 1373 K and a pressure range of 4-7 atm.
Kinetic models of n-heptane oxidation have been proposed (e.g. [28-31]) and
validated over a wide range of experimental conditions. Glaude et al. [32] and Battin-
Leclerc et al. [33] carried out a detailed modeling of the oxidation of n-octane and n-
decane. Zeppieri et al. [34] developed a skeletal kinetic mechanism for high and
intermediate temperature n-decane oxidation, which has been modeled also by Bikas et al.
[35] and validated for a wide range of combustion regimes. Ranzi et al. [36] proposed a
lumped model for studying the oxidation of heavy n-alkanes from n-decane to n-
hexadecane, and demonstrated that heavy n-alkanes display the same kinetic behavior
both at low-temperatures as well as high-temperatures. You et al. [37] proposed a high-
temperature detailed kinetic model for the combustion of n-alkanes up to n-dodecane. In
order to provide insight into the effect of molecular size on ignition, Westbrook et al. [38]
developed recently a detailed kinetic model describing the oxidation of n-alkanes from n-
octane to n-hexadecane, and showed that the ignition delay time increases as the length of
the n-alkane chain decreases at post-shock temperatures just below 1000 K.
In practical combustion devices ignition is obtained in the presence of temperature
and species concentration gradients so that transport processes could affect positively or
50
negatively the evolution of the ignition kinetics. The counterflow configuration provides
an ideal environment for flame ignition studies. As a result it has been used widely in the
recent past for both premixed (e.g. [39,40]) and non-premixed (e.g. [41-48]) flames of
hydrogen, CO, and C
1
-C
4
n-alkanes.
Hydrogen addition [40,43] has been found to improve hydrocarbon flame ignition
significantly. Methane flames have been determined to exhibit higher flame ignition
temperatures compared to C
2
-C
4
n-alkanes due to the effective H-radical termination
caused by the methyl radicals [44]. Varying the molecular size from ethane to n-butane,
it has been shown that the ignition temperature of non-premixed flames increases with
the molecular size [46,48] due to the attendant decrease of the fuel diffusivity.
There are relatively fewer flame ignition studies for n-alkanes heavier than C
5
, with
the exception being n-heptane [49-51]. Seiser et al. [49] performed experimental and
numerical studies of autoignition of n-heptane in counterflow non-premixed flames, and
found that high-temperature kinetics control autoignition while low-temperature kinetics
becomes gradually more important with decreasing strain rate. The ignition temperature
of n-heptane counterflow flames was found to be very sensitive to its diffusion rate [50].
Blouch and Law [52] determined the ignition temperatures of non-premixed flames of
C
1
-C
7
n-alkanes at atmospheric pressure, and a distinctly monotonic increase of the
ignition temperature with the fuel size was not realized. Holley et al. [53] studied the
ignition of C
5
-C
14
n-alkane non-premixed flames by counterflowing a fuel/N
2
jet against
vitiated air resulting from ultra-lean H
2
/CO/air flames. Results showed that the lower is
the carbon number the greater is the ignition propensity. Seshadri et al. [54] determined
51
ignition temperatures of non-premixed C
7
-C
16
flames using the liquid pool variation of
the counterflow configuration, and it was reported that the ignition temperature decreases
with increasing fuel size. Liu et al. [51] performed an experimental study on flame
ignition of a liquid pool of n-heptane as well. The effect of reactant diffusion on non-
premixed flame ignition has been found computationally to be notable, especially for
cases in which the fuel is diluted highly by an inert [55].
Considering the ignition studies for n-alkanes in both homogeneous systems and
flames, it is apparent that as the fuel molecular weight increases there are two competing
mechanisms. On one hand, the fuel reactivity increases with the molecular size
facilitating thus ignition. On the other hand and in the presence of species concentration
gradients, the fuel diffusive transport decreases retarding thus ignition, although this has
not been realized in two earlier flames studies [52,53].
Based on these considerations, the main goal of the present study was to provide
archival ignition data for non-premixed liquid n-alkane flames for the wide C
5
-C
12
carbon
range, and to gain insight into the underlying physical and chemical mechanisms with the
aid of detailed numerical simulations using the recently developed JetSurF (version 1.0)
kinetic model [56]. Propane flames were considered also, being the simplest gaseous n-
alkane whose oxidation characteristics mimic those of the large molecular weight ones.
5.2 Experimental Approach
Experiments were carried out under atmospheric pressure in the counterflow
configuration, as schematically shown in Figure 2.1. Burners with 22 mm diameter
52
nozzles were used, and the nozzle separation distance was 20 mm. T
u
= 448 K for all C
3
-
C
12
n-alkanes considered in this Chapter. In all experiments, the exit velocities were
115 cm/s for the fuel-containing jet and 256 cm/s for the oxidizer jet. Under such
conditions, the local strain rate was measured and it was determined to be nearly constant,
that is K = 140 s
-1
.
Measurements were performed for mixtures of N
2
with C
3
H
8
(100%), n-C
5
H
12
(Alfa
Aesar, 98%), n-C
6
H
14
(Alfa Aesar, 99+%), n-C
7
H
16
(Alfa Aesar, 99%), n-C
8
H
18
(Alfa
Aesar, 98%), n-C
9
H
20
(Alfa Aesar, 99%), n-C
10
H
22
(TCI America, 99%), and n-C
12
H
26
(TCI America, 99%). Most of the impurities are isomers of similar molecular weight. In
all reagents used, the water contents are less than 0.01%. n-C
11
H
24
was not considered in
the present study as analysis has shown [37] that its oxidation kinetics does not play any
major role to those of n-C
12
H
26
and the lower carbon number n-alkanes.
In this Chapter, the fuel mole fraction range in the fuel/N
2
mixture, X
F
, was varied
from 1% to 12%.
5.3 Modeling Approach
T
ign
was computed using an opposed-jet flow code. Both the mixture-averaged and
full multi-component transport coefficient formulations were used in the simulations
along with the Soret effect and thermal radiation.
The local strain rate in all simulations were matched with the value in experiments
that was K = 140 s
-1
. For the conditions considered in the present Chapter the measured
velocity gradient at the burner exits was 40 s.
-1
53
The recently developed kinetic model JetSurF 1.0 [56] was adopted in the simulations.
The model consists of 194 species and 1459 reactions, which describes the pyrolysis and
high-temperature kinetics of normal alkanes up to n-dodecane.
5.4 Results and Discussion
5.4.1 Ignition of n-dodecane Flames
The effect of fuel concentration on ignition can be assessed by plotting T
ign
versus X
F
,
on a percentage basis. For example, Figure 5.1 depicts the experimental and computed
T
ign
for n-C
12
H
26
flames. The experimental data show that T
ign
monotonically decreases
with X
F
and this is more profound for low values of X
F
. This behavior is in agreement
with that of hydrogen and C
1
-C
4
hydrocarbons as reported by Fotache et al. [43,44,46]
and n-heptane as reported by Liu et al. [51]. At low X
F
, flame ignition is largely
diffusion-controlled and as a result there is a notable dependence of T
ign
on X
F
. As X
F
increases, this dependence becomes more gradual, which is physically sound given that
the effect of diffusion diminishes and thus the dependence of T
ign
on X
F
. The simulations
using the multi-component transport coefficient formulation agree very closely with
experimental data. The computed T
ign
using the mixture-averaged transport coefficient
formulation are shown also in Figure 5.1. The discrepancies between the two
formulations vary between 16 K and 10 K for high and low T
ign
respectively. With the
exception of the results shown in Figure 5.1, all computed T
ign
’s that are reported in this
Chapter, were determined using the multi-component transport coefficient formulation.
54
Figure 5.1 Experimental and computed T
ign
of n-C
12
H
26
flames; ( ─) Simulation
results using the JetSurF 1.0 reaction model with multi-component
transport coefficient formulation; ( ┄) Simulation results with
mixture-averaged transport coefficient formulation.
The ranked LSC’s of T
ign
on reaction rate coefficients and BDC’s are shown in
Figures 5.2 and 5.3 respectively, for n-C
12
H
26
flames with X
F
= 2% and 10%. As shown
in Figure 5.2, T
ign
is sensitive mainly to the main branching reaction R1, the termination
reaction R2, the CO oxidation reaction R3, the OH-consuming reaction R4, the formyl
radical (HCO) reactions R5 and R6, the vinyl radical (C
2
H
3
) reactions R7-R10, and the
methyl radical (CH
3
) reaction R11. The importance of R1 in promoting ignition is
apparent given that it provides the necessary chain branching. R3 and R5 are sources of
H radicals required by R1, while R3 is the main CO oxidation pathway promoting heat
release and H production. R2 and R6 inhibit ignition as they compete directly for
reactants with R1 and R5 resulting in essence in chain-termination. In fact, at lower X
F
,
1150
1200
1250
1300
1350
1400
1450
123456789 10 11
Ignition Temperature, T
ign
, K
Fuel Mole Fraction, X
F
,%
55
ignition becomes more sensitive to reactions R1, R2, R3, R5, and R6. R7 is promoting
ignition because it produces O and CH
2
CHO that is consumed via R12 and R13 to
produce CH
3
and HO
2
. Reaction path analysis indicates that CH
3
and HO
2
are primarily
consumed through R11, which results in OH and CH
3
O that can decompose readily to
CH
2
O and H promoting thus ignition. R8 and R9 promote ignition as well, being sources
of C
2
H
3
, while R10 inhibits ignition as it competes directly for reactants with R7 to form
relatively more stable species. Reactions involving n-C
12
H
26
and other large hydrocarbon
intermediates are generally too fast to be rate limiting [57], and as a result no measurable
sensitivity was determined.
H + O
2
→ O + OH (R1)
H + O
2
+ M → HO
2
+ M (R2)
CO + OH → CO
2
+ H (R3)
OH + HO
2
→ H
2
O + O
2
(R4)
HCO + M → CO + H + M (R5)
HCO + O
2
→ CO + HO
2
(R6)
C
2
H
3
+ O
2
→ CH
2
CHO + O (R7)
C
2
H
4
+ OH → C
2
H
3
+ H
2
O (R8)
aC
3
H
5
+ HO
2
→ OH + C
2
H
3
+ CH
2
O (R9)
C
2
H
3
+ O
2
→ HCO + CH
2
O (R10)
CH
3
+ HO
2
→ CH
3
O + OH (R11)
CH
2
CHO → CH
3
+ CO (R12)
CH
2
CHO + O
2
→ CH
2
CO + HO
2
(R13)
56
Figure 5.2 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients, computed for n-C
12
H
26
flames at X
F
= 2% and
X
F
= 10%.
Figure 5.3 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for n-C
12
H
26
flames at X
F
= 2% and X
F
= 10%.
-5.00E-02 -4.00E-02 -3.00E-02 -2.00E-02 -1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
HCO+O
2
→CO+HO
2
aC
3
H
5
+HO
2
→OH+C
2
H
3
+CH
2
O
C
2
H
4
+OH →C
2
H
3
+H
2
O
C
2
H
3
+O
2
→CH
2
CHO+O
CH
3
+HO
2
→CH
3
O+OH
HCO+M →CO+H+M
CO+OH →CO
2
+H
H+O
2
→O+OH
C
2
H
3
+O
2
→HCO+CH
2
O
OH+HO
2
→H
2
O+O
2
H+O
2
+M →HO
2
+M
aC
3
H
5
+H+M →C
3
H
6
+M
C
3
H
6
+OH →aC
3
H
5
+H
2
O
X
F
=2%
X
F
=10%
-1.00E-01 -8.00E-02 -6.00E-02 -4.00E-02 -2.00E-02 0.00E+00 2.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
CH
2
O—N
2
OH—N
2
H
2
—N
2
CO—N
2
C
2
H
4
—N
2
C
3
H
6
—N
2
O
2
—N
2
O
2
—n -C
12
H
26
n -C
12
H
26
—N
2
X
F
=2%
X
F
=10%
57
As shown in Figure 5.3, LSC’s of T
ign
on the n-C
12
H
26
-N
2
BDC are negative and
approximately twice as large in magnitude compared to the attendant sensitivity
coefficient of the main branching reaction R1. Hence, the ignition process for the
conditions considered herein is limited by the fuel diffusive transport towards the ignition
kernel, which is defined as the region where the concentrations of radicals peak (e.g.,
[43,44,46]) and is located on the oxidizer side of the stagnation plane, i.e. closer to the
highest temperature in the flow field. In comparison, the authors have shown [57] that
the propagation of n-C
12
H
26
flames is rather sensitive to the O
2
-N
2
BDC and minimally to
the n-C
12
H
26
-N
2
BDC. It can be seen also that compared to X
F
= 10%, T
ign
is slightly
more sensitive to fuel diffusion at X
F
= 2%, which explains quantitatively the change of
the T
ign
.vs . X
F
slope as X
F
increases.
5.4.2 Ignition of C
3
– C
10
n-alkane Flames
The experimental and computed T
ign
’s for C
3
and C
5
-C
10
n-alkanes are shown in
Figures 5.4 and 5.5, along with computed results that are in close agreement with the
experimental data, with the exception being the slight over-prediction of the C
3
H
8
data.
Figure 5
(a)
(c)
5.4 Experi
C
6
H
14
,
simula
compo
imental and
and (d) n-C
ation results
onent transpo
computed
C
7
H
16
flames;
using the Je
ort coefficien
T
ign
’s; (a) C
; Symbols: e
etSurF 1.0 k
nt formulatio
(b)
(d)
C
3
H
8
, (b) n-
experimental
kinetic mode
on.
-C
5
H
12
, (c) n
l data; Lines
el with multi
58
n-
s:
i-
F
fo
re
in
Figure 5
LSC’s of
igures 5.6 an
or n-C
12
H
26
eveal that th
nvolving the
(a)
(c)
5.5 Experi
n-C
10
H
results
transp
f T
ign
on reac
nd 5.7 respe
flames show
he controllin
iC
3
H
7
radic
imental and
H
22
flame; S
s using the J
ort coefficien
ction rate co
ectively, at X
wn in Figur
ng reactions
cal appear to
computed T
Symbols: ex
JetSurF 1.0
nt formulatio
oefficients an
X
F
= 3% and
res 5.2 and
s are simila
o be importan
T
ign
’s; (a) n-C
xperimental
kinetic mod
on.
nd BDC’s fo
10%. Com
5.3, the res
ar in genera
nt also as it
(b)
C
8
H
18
, (b) n-C
data; Lines
del with mul
for C
3
H
8
flam
mpared to the
sults of Fig
al. Reaction
is produced
C
9
H
20
, and (c
s: simulatio
lti-componen
mes are show
e attendant re
gures 5.6 an
ns R14 and
in large am
59
c)
on
nt
wn in
esults
d 5.7
d R15
ounts
d
co
irectly from
ompetes wit
iC
3
H
7
+ M
iC
3
H
7
+ O
Figure 5
m C
3
H
8
. R1
h R14 for iC
M → C
3
H
6
+
O
2
→ C
3
H
6
+
5.6 Logari
coeffic
14 is produc
C
3
H
7
and exh
+ H + M
+ HO
2
ithmic sens
ients, compu
cing H radic
hibits thus an
sitivity coef
uted for C
3
H
8
cals promot
n inhibitive e
fficients of
8
flames at X
ing thus ign
effect on ign
T
ign
on r
X
F
= 3% and X
nition while
nition.
(R
(R
reaction rat
X
F
= 10%.
60
e R15
R14)
R15)
te
61
Figure 5.7 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for C
3
H
8
flames at X
F
= 3% and X
F
= 10%.
Comparing the results of Figures 5.6 and 5.7, it can be seen that the LSC value of T
ign
on the C
3
H
8
–N
2
BDC is more than 2.5 times greater than the highest sensitivity on
kinetics, indicating that the ignition process for the conditions considered herein is still
limited by the fuel diffusivity even for the smallest n-alkane considered in the present
study.
-1.00E-01 -8.00E-02 -6.00E-02 -4.00E-02 -2.00E-02 0.00E+00 2.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
CH
2
O—N
2
OH—N
2
H
2
—N
2
CO—N
2
C
2
H
4
—N
2
C
3
H
6
—N
2
O
2
—N
2
O
2
—C
3
H
8
C
3
H
8
—N
2
X
F
=3%
X
F
=10%
5
ef
an
X
as
ar
.4.3 Comp
Figure 5
Since the
ffects on ign
nd computed
X
F
; X
F
was c
ssessing the
re determin
parisons of t
5.8 Compa
( ◇)( ─
Symbo
the Je
coeffic
fuel diffusi
nition are ass
d T
ign
’s of n
chosen over
competition
ned by the
the Ignition
arison of ex
─) n-C
5
H
12
,
ols: present e
etSurF 1.0
ient formula
ivity decreas
sessed via co
n-C
5
H
12
, n-C
the corresp
n between th
e rate of m
n of C
3
– C
12
perimentally
( ■)( …) n-C
experimental
kinetic mod
ation.
ses as the m
omparisons
C
7
H
16
, and n-
ponding mas
he chemical
molecular c
2
n-alkane F
y and numer
C
7
H
16
, and (
l data; Lines
del with mu
molecular we
in Figure 5.
-C
12
H
26
flam
ss fraction, a
l kinetics an
collisions,
Flames
rically deter
( ▲)( ┄) n-C
s: simulation
ulti-compone
eight increas
.8, in which
mes are show
as it is mor
nd fuel diffu
in other w
rmined T
ign
’s
C
12
H
26
flames
n results usin
ent transpor
ses, the atten
the experim
wn as functi
e appropriat
usion whose
words on m
62
s;
s.
ng
rt
ndant
mental
on of
te for
rates
molar
63
concentrations. It is of particular interest to note that for the larger values of X
F
considered in this study, the experimental and computed T
ign
values are higher for the
heavier fuel the opposite is true for the lower values of X
F
. More specifically, the
T
ign
.vs . X
F
curves for the three fuels cross at X
F
< 4%, which appears to be counter-
intuitive based on physical arguments.
Sensitivity analyses of T
ign
on kinetics and BDC’s were performed for n-C
5
H
12
, n-
C
7
H
16
, and n-C
12
H
26
flames at X
F
= 2% and 10% and the results are shown in Figures 5.9
and Figures 5.10 respectively. Reactions R1-R11 involving H, CO, HCO, CH
3
, C
2
H
3
,
C
2
H
4
and C
3
H
6
dominate ignition, while R1 and R6 exhibit the largest negative and
positive LSC respectively, for all the three fuels. As the fuel size decreases, T
ign
becomes
more sensitive to kinetics for both X
F
= 2% and 10%. Note that the importance of R11
increases at X
F
= 10% because at higher fuel concentrations more CH
3
is produced
promoting thus the rate of R11.
As shown in Figures 5.10, LSC’s of T
ign
on the fuel-N
2
BDC is still the largest among
all the species pairs. It is worthwhile to note that the sensitivities to the C
2
H
4
-N
2
and
C
3
H
6
-N
2
BDC’s are noticeable as well. C
2
H
4
and C
3
H
6
are major n-alkene products of
the rapid n-alkane -scission that is accelerated as temperature increases. It can be seen
that for X
F
= 2% that corresponds to higher T
ign
, the sensitivity to the C
2
H
4
-N
2
BDC
increases confirming that the fate of the products of the fuel decomposition needs to be
considered also in assessing the ignition response.
64
(a)
(b)
Figure 5.9 Logarithmic sensitivity coefficients of T
ign
on reaction rate
coefficients computed for n-C
5
H
12
, n-C
7
H
16
, and n-C
12
H
26
flames; (a)
X
F
= 2%, (b) X
F
= 10%.
-5.00E-02 -4.00E-02 -3.00E-02 -2.00E-02 -1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
HCO+O
2
→CO+HO
2
aC
3
H
5
+HO
2
→OH+C
2
H
3
+CH
2
O
C
2
H
4
+OH →C
2
H
3
+H
2
O
C
2
H
3
+O
2
→CH
2
CHO+O
CH
3
+HO
2
→CH
3
O+OH
HCO+M →CO+H+M
CO+OH →CO
2
+H
H+O
2
→O+OH
C
2
H
3
+O
2
→HCO+CH
2
O
OH+HO
2
→H
2
O+O
2
H+O
2
+M →HO
2
+M
C
3
H
6
+OH →aC
3
H
5
+H
2
O
n -C
5
H
12
n -C
7
H
16
n -C
12
H
26
-5.00E-02 -4.00E-02 -3.00E-02 -2.00E-02 -1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
HCO+O
2
→CO+HO
2
aC
3
H
5
+HO
2
→OH+C
2
H
3
+CH
2
O
C
2
H
4
+OH →C
2
H
3
+H
2
O
C
2
H
3
+O
2
→CH
2
CHO+O
CH
3
+HO
2
→CH
3
O+OH
HCO+M →CO+H+M
CO+OH →CO
2
+H
H+O
2
→O+OH
C
2
H
3
+O
2
→HCO+CH
2
O
OH+HO
2
→H
2
O+O
2
H+O
2
+M →HO
2
+M
C
3
H
6
+OH →aC
3
H
5
+H
2
O
n -C
5
H
12
n -C
7
H
16
n -C
12
H
26
65
The structure of the ignition kernel was analyzed for n-C
5
H
12
and n-C
12
H
26
flames
just prior to ignition. The concentration profiles of C
2
H
4
, n-C
5
H
12
, and n-C
12
H
26
for
X
F
= 2% with T
air
= 1333 K and for X
F
= 10% with T
air
= 1226 K are shown in Figures
5.11a and 5.11b respectively. Ignition under these conditions is generally controlled by
the diffusive transport of reactants towards the ignition kernel. It is of interest to note
that for both n-C
5
H
12
and n-C
12
H
26
, the fuel is largely consumed upon reaching the
ignition kernel. Thus, it is essential to account also for the processes controlling the fate
of the immediate products of fuel decomposition, such as C
2
H
4
. It was determined
through reaction path analysis, that C
2
H
4
is produced mainly via reaction R16 that is also
the primary production pathway of H. Additionally, the analysis revealed that C
2
H
5
is
produced through a sequence of reactions stemming from the rapid -scission of the fuel,
such as the primary consumption pathways of n-C
5
H
12
or n-C
12
H
26
noted below as P1 and
P2 respectively, which is more profound for n-C
12
H
26
compared to n-C
5
H
12
at the higher
temperatures encountered at X
F
= 2%.
C
2
H
5
+ M → C
2
H
4
+ H + M (R16)
n-C
5
H
12
→ C
5
H
11
→ C
2
H
5
→ C
2
H
4
(P1)
n-C
12
H
26
→ C
12
H
25
→ C
4
H
9
→ C
2
H
5
→ C
2
H
4
(P2)
66
(a)
(b)
Figure 5.10 Logarithmic sensitivity coefficients of T
ign
on binary diffusion
coefficients computed for n-C
5
H
12
, n-C
7
H
16
, and n-C
12
H
26
flames;
(a) X
F
= 2%, (b) X
F
= 10%; the logarithmic sensitivity coefficient on
the main branching reaction is shown for comparison purposes.
-1.00E-01 -8.00E-02 -6.00E-02 -4.00E-02 -2.00E-02 0.00E+00 2.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
CH
2
O—N
2
OH—N
2
H
2
—N
2
CO—N
2
C
2
H
4
—N
2
C
3
H
6
—N
2
O
2
—N
2
O
2
—fuel
fuel—N
2
O
2
—C
2
H
4
H
2
O—N
2
H—N
2
n -C
5
H
12
n -C
7
H
16
n -C
12
H
26
H+O
2
→O+OH
-1.00E-01 -8.00E-02 -6.00E-02 -4.00E-02 -2.00E-02 0.00E+00 2.00E-02
Logarithmic Sensitivity Coefficients of Ignition Temperature
CH
2
O—N
2
OH—N
2
H
2
—N
2
CO—N
2
C
2
H
4
—N
2
C
3
H
6
—N
2
O
2
—N
2
O
2
—fuel
fuel—N
2
HO
2
—N
2
H+O
2
→O+OH
n -C
5
H
12
n -C
7
H
16
n -C
12
H
26
Figure 5
5.11 Conce
X
F
= 2
C
5
H
12
entration pr
2%, T
air
= 13
2
flames; ( ─)
(a)
(b)
rofiles of C
333 K and (
) n-C
12
H
26
fla
C
2
H
4
, n-C
5
H
(b) X
F
= 10%
ames.
H
12
, and n-C
%, T
air
= 122
C
12
H
26
at (a
26 K; ( ┄) n
67
a)
n-
68
As shown in Figure 5.11a for n-C
12
H
26
flames at X
F
= 2%, the concentration of C
2
H
4
in the ignition kernel is greater compared to n-C
5
H
12
and this behavior is reversed at
X
F
= 10% shown in Figure 5.11b. Given that the ignition propensity of n-alkenes is
higher compared to n-alkanes, especially for C
2
H
4
(e.g., [58]), and that, in general, the
products of fuel decomposition are more diffusive compared to the fuel, at X
F
= 2% n-
C
12
H
26
flames ignite more readily, i.e. exhibit lower T
ign
than n-C
5
H
10
flames, and the
opposite is true at X
F
= 10%. There is a large difference of the order of 100 K between
the T
ign
values at X
F
= 2% and 10%, resulting thus in more extensive n-C
12
H
26
decomposition compared to n-C
5
H
12
at X
F
= 2% and the production of higher amounts of
C
2
H
4
.
The concentration profiles of the highly active H and OH radicals are shown in
Figures 5.12a and 5.12b for conditions that are identical to those reported in Figures
4.11a and 4.11b. At X
F
= 2% and T
air
= 1346 K (Figure 5.12a), the H and OH
concentrations for n-C
12
H
26
flames are distinctly higher compared to n-C
5
H
12
, while the
opposite behavior is observed at X
F
= 10% and T
air
= 1258 K (Figure 5.12b).
Figure 5
5.12 Conce
X
F
= 2
C
5
H
12
entration pr
2%, T
air
= 13
2
flames; ( ─)
(a)
(b)
rofiles of H,
333 K and (
) n-C
12
H
26
fla
, OH, n-C
5
H
(b) X
F
= 10%
ames.
H
12
, and n-C
%, T
air
= 122
C
12
H
26
at (a
26 K; ( ┄) n
69
a)
n-
70
Given that H, OH, and C
2
H
4
are known to promote ignition, the results of Figures
5.11 and 5.12 provide insight into the mechanisms causing the non-monotonic
dependence of T
ign
on the fuel molecular weight, as X
F
varies. More specifically, as X
F
decreases, T
ign
increases resulting thus in more extensive fuel decomposition, and the
more diffusive and reactive products of decomposition influence the ignition behavior.
On the other hand, as X
F
increases, T
ign
decreases so that the fuel decomposition is
minimum and as a result, the ignition response is controlled largely by the fuel transport
and kinetics.
The experimentally determined T
ign
’s of all C
3
-C
12
n-alkane flames are compared and
shown in Figure 5.13 (a) and (b) as function of X
F
and Y
F
, respectively. In general, T
ign
of
the heavier n-alkane flames exhibits a milder response to X
F
compared to lighter n-
alkanes. It can be seen that for large (low) X
F
, the experimental T
ign
is higher (lower) for
the heavier fuel. Thus, a “crossing” of the T
ign
responses is observed at low X
F
when
comparing experimentally determined T
ign
’s for any two n-alkanes. That is the result of
the competition between the fuel reactivity and diffusive transport.
Sensitivity analysis of T
ign
on reaction rates and BDC’s for all C
5
-C
10
n-alkane flames
revealed results that are very similar to n-C
12
H
26
flames. Reaction path analysis showed
that the fuel is consumed largely by H-abstractions and the resulting alkyl radicals are
subsequently decomposed through rapid -scission to n-alkene products, such as C
3
H
6
and C
2
H
4
, prior to ignition.
Figure 5 5.13 Comp
C
5
H
12
C
10
H
2
Inset:
Data
parison of ex
2
, ( ♦) n-C
6
H
1
22
, and ( ▲)
: detail of th
based on fue
(a)
(b)
xperimentall
14
, ( ■) n-C
7
H
n-C
12
H
26
. (a
he region of
el mass fracti
ly determine
H
16
, ( □) n-C
8
a) Data base
f fuel mole f
ion.
ed T
ign
’s; ( +)
8
H
18
, ( ∆) n-C
ed on fuel m
fractions bel
) C
3
H
8
, ( *) n
C
9
H
20
, ( ◇) n
mole fraction
low 4%; (b
71
n-
n-
n,
b)
72
5.5 Concluding Remarks
Ignition temperatures of non-premixed C
3
-C
12
n-alkane flames were determined
experimentally in the counterflow configuration under atmospheric pressure, a constant
strain rate of 140s
-1
determined on the fuel size, and at an unburned fuel/N
2
mixture
temperature of 448 K. The flow velocities were determined using Laser Doppler
Velocimetry and the reported strain rate was measured locally. A recently developed
detailed kinetic model (JetSurF version 1.0), consisting of 194 species and 1459 reactions,
was used in the numerical simulations of the experimental data. It was determined also
that there is a notable discrepancy in the predicted ignition temperatures when using
multi-component transport coefficient formulation compared to the computationally less
intense mixture-averaged formulation.
Experimental results revealed that the ignition temperature decreases monotonically
with increasing fuel concentration and such dependence is more profound for low fuel
concentrations for all C
3
-C
12
n-alkanes considered in this study. Furthermore, both the
experimental and computed results indicate that the ignition temperature of heavier n-
alkanes exhibit a milder dependence on the fuel concentration compared to lighter n-
alkanes. The computed ignition temperatures were found to be in good agreement, with
the exception of the propane flames for which the data were over-predicted slightly.
Detailed sensitivity analyses on kinetics and molecular transport revealed that for the
conditions considered in this study, ignition is sensitive to both fuel diffusion and the
kinetics of low molecular weight intermediates. The dependence of the ignition
temperature on the fuel molecular size was found to be different at low and high fuel
73
concentrations. This behavior was attributed to the competition between fuel reactivity
and diffusivity, and the extent if this competition was determined to depend on the fuel
size.
5.6 References
[1] A. Burcat, K. Scheller, A. Lifshitz, Combust. Flame 16 (1971) 29-33.
[2] J. Franck, J.F. Griffiths, W. Nimmo, Proc. Combust. Inst. 21 (1988) 447–454.
[3] M. Carlier, C. Corre, R. Minetti, J.F. Pauwels, M. Ribaucour, L.R. Sochet, Proc.
Combust. Inst. 23 (1991) 1753–1758.
[4] S. Kojima, T. Suzuoki, Combust. Flame 92 (1993) 254–265.
[5] S. Kojima, Combust. Flame 99 (1994) 87–136.
[6] R. Minetti, M. Ribaucour, M. Carlier, C. Fittschen, L.R. Sochet, Combust. Flame
96 (1994) 201–211.
[7] R. Minetti, M. Ribaucour, M. Carlier, L.R. Sochet, Combust. Sci. Technol. 113–
114 (1996) 179–192.
[8] J.F. Griffiths, P.A. Halford-Maw, D.J. Rose, Combust. Flame 95 (1993) 291–306.
[9] H. Kim, Y. Lim, K. Min, D. Lee, KSME Int. J. 16 (8) (2002) 1127–1134.
[10] D. Healy, N.S. Donato, C.J. Aul, E.L. Petersen, C.M. Zinner, G. Bourque, H.J.
Curran, Combust. Flame 157 (2010) 1526-1539.
[11] S. Gersen, A.V. Mokhov, J.H. Darmeveil, H.B. Levinsky, Combust. Flame 157
(2010) 240-245.
74
[12] D.C. Horning, D.F. Davidson, R.K. Hanson, J. Propul. Power 18 (2) (2002) 363–
371.
[13] H.K. Ciezki, G. Adomeit, Combust. Flame 93 (1993) 421-433.
[14] R. Minetti, M. Carlier, M. Ribaucour, E. Therssen, L.R. Sochet, Combust. Flame
102 (1995) 298-309.
[15] D.J. Vermeer, J.W. Meyer, A.K. Oppenheim, Comb. Flame 18 (1972) 327-336.
[16] C.M. Coats, A. Williams, Symposium (International) on Combustion, 17 (1979)
611–621.
[17] K. Fieweger, R. Blumenthal, G. Adomeit, Proc. Combust. Inst. 25 (1994) 1579-
1585.
[18] A. Cox, J.F. Griffiths, C. Mohamed, H.J. Curran, W.J. Pitz, C.K. Westbrook,
Twenty-Sixth Symposium (International) on Combustion, The Combustion
Institute, (1996) pp. 2685–2692.
[19] M.B. Colket, III, L.J. Spadaccini, J Propul Power 17 (2001) 315-323.
[20] B.M. Gauthier, D.F. Davidson, R.K. Hanson, Comb. Flame 139 (2004) 300-311.
[21] D.F. Davidson, M.A. Oehlschlaeger, R.K. Hanson, Proc. Combust. Inst. 31 (2007)
321-328.
[22] E. Olchanski, A. Burcat, Int. J. Chem. Kinet. 38 (2006) 703-713.
[23] V.P. Zhukov, V.A. Sechenov, A.Y. Starikovskii, Combust. Flame 153 (2008) 130-
136.
[24] K. Kumar, G. Mittal, C.-J. Sung, Combust. Flame 156 (2009) 1278-1288.
75
[25] S.S. Vasu, D.F. Davidson, Z. Hong, V. Vasudevan, R.K. Hanson, Proc. Combust.
Inst. 32 (2009) 173-180.
[26] D.F. Davidson, Z. Hong, G.L. Pilla, A. Farooq, R.D. Cook, R.K. Hanson, Proc.
Combust. Inst. 33 (2011) 151-157.
[27] D.R. Haylett, R.D. Cook, D.F. Davidson, R.K. Hanson, Proc. Combust. Inst. 33
(2011) 167-173.
[28] E. Ranzi, P. Gaffuri, T. Faravelli, P. Dagaut, Combust. Flame 103 (1995) 91-106.
[29] R.P Lindstedt, L.Q. Maurice, Combust. Sci. Technol. 107 (1995) 317-353.
[30] T.J. Held, A.J. Marchese, F.L. Dryer, Combust Sci Technol 123 (1997) 107-146.
[31] H.J. Curran, P. Gaffuri, W.J. Pitz, C.K. Westbrook, Comb. Flame 114 (1998) 149-
177.
[32] P.A. Glaude, V. Warth, R. Fournet, F. Battin-Leclerc, G. Scacchi, G.M. Côme, Int
J Chem Kinet 30 (1998) 949–959.
[33] F. Battin-Leclerc, R. Fournet, P.A. Glaude, B. Judenherc, V. Warth, G.M. Côme,
G. Scacchi, Proc Combust Inst 28 (2000) 1597-1605.
[34] S.P. Zeppieri, S.D. Klotz, F.L. Dryer, Proc. Combust. Inst. 28 (2000) 1587-1595.
[35] G. Bikas, N. Peters, Combust Flame 126 (2001) 1456-1475.
[36] E. Ranzi, A. Frassoldati, S. Granata, T. Faravelli, Ind. Eng. Chem. Res. 44 (2005)
5170–5183.
[37] X. You, F.N. Egolfopoulos, H. Wang, Proc. Combust. Inst. 32 (2009) 403-410.
[38] C.K. Westbrook, W.J. Pitz, O. Herbinet, H.J. Curran, E.J. Silke, Combust. Flame
156 (2009) 181-199.
76
[39] X.L. Zheng, C.K. Law, Combust. Flame 136 (2004) 168-179.
[40] O. Park, P.S. Veloo, N. Liu, F.N. Egolfopoulos, Proc. Combust. Inst. 33 (2011)
887-894.
[41] T.G. Kreutz, M. Nishioka, C.K. Law, Combust. Flame 99 (1994) 758-766.
[42] C.G. Fotache, T.G. Kreutz, D.L. Zhu, C. K. Law, Combust. Sci. Technol. 109
(1995) 373–393.
[43] C.G. Fotache, T.G. Kreutz, C. K. Law, Combust. Flame 110 (1997) 429-440.
[44] C.G.Fotache, T.G. Kreutz, C. K. Law, Combust. Flame 108 (1997) 442-470.
[45] T.G. Kreutz, C.K. Law, Combust. Flame 114 (1998) 436-456.
[46] C.G. Fotache, H. Wang, C. K. Law, Combust. Flame 117 (1999) 777-794.
[47] C.G. Fotache, Y. Tan, C.J. Sung, C. K. Law, Combust. Flame 120 (2000) 417-426.
[48] S. Humer, R. Seiser, K. Seshadri, Proc. Combust. Inst. 29 (2002) 1597-1604.
[49] R. Seiser, H. Pitsch, K. Seshadri, W.J. Pitz, H.J. Curran, Proc. Combust. Inst. 28
(2000) 2029-2037.
[50] A.J. Smallbone, W. Liu, C.K. Law, X.Q. You, H. Wang, Proc. Combust. Inst. 32
(2009) 1245-1252.
[51] W. Liu, D. Zhu, N. Wu, C.K. Law, Combust. Flame 157 (2010) 259-266.
[52] J.D. Blouch, C.K. Law, Proc. Combust. Inst. 28 (2000) 1679-1686.
[53] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, T. Edwards, Proc.
Combust. Inst. 31 (2007) 1205-1213.
[54] K. Seshadri, S. Humer, R. Seiser, Combust. Theor. Model. 12 (2008) 831-855.
[55] M.G. Andac, F.N. Egolfopoulos, Proc. Combust. Inst. 31 (2007) 1165-1172.
77
[56] B. Sirjean, E. Dames, D.A. Sheen, X.-Q. You, C. Sung, A.T. Holley, F.N.
Egolfopoulos, H. Wang, S.S. Vasu, D.F. Davidson, R.K. Hanson, H. Pitsch, C.T.
Bowman, A. Kelley, C.K. Law, W. Tsang, N.P. Cernansky, D.L. Miller, A. Violi,
R.P. Lindstedt, A high-temperature chemical kinetic model of n-alkane oxidation,
JetSurF version 1.0, September 15, 2009
(http://melchior.usc.edu/JetSurF/Version1_0/Index.html).
[57] C. Ji, E. Dames, Y.L. Wang, H. Wang, F.N. Egolfopoulos, Combust. Flame 157
(2010) 277–287.
[58] W. Liu, A.P. Kelley, C.K. Law, Comb. Flame 157 (2010) 1027-1036.
78
Chapter 6
Ignition of Non-Premixed Counterflow Flames of Octane and Decane
Isomers
6.1 Introduction
Ignition is of prime importance in combustion engines (e.g., [1]). Branched
hydrocarbons are known to have reduced ignition and as a result knocking propensity,
and they are present in notable quantities in gasoline (e.g., [2]). Additionally, branched
hydrocarbons are present in conventional and synthetic diesel and jet fuels.
The combustion characteristics of branched alkanes have been studied to lesser extent
compared to n-alkanes, with the exception of iso-octane (2,2,4-trimethyl-pentane) that is
a key component of primary reference fuels (PRF), as well as a representative of iso-
alkanes in jet fuel surrogates. Numerous experimental (e.g., [3-14]) and modeling (e.g.,
[15-18]) studies have been performed on the ignition kinetics of iso-octane in flames and
homogeneous systems.
Studies of isomeric C
4
-C
8
hydrocarbons have attracted considerable attention in the
past and recent years (e.g., [19-25]). Salooja [19,20] investigated the ignition
79
characteristics of the hexane isomers under 893K at atmospheric pressure in a quartz
reaction chamber and reported that increasing the degree of branching lowers the
tendency towards oxidation inhibiting ignition, and that H
2
O
2
and HCHO are important
chain branching intermediates approaching to ignition. Ribaucour et al. [21] examined
the low-temperature ignition kinetics of three isomers of pentane in a rapid compression
machine (RCM). Silke et al. [22] measured ignition delay times of nine isomers of
heptane in a RCM for a gas temperature range of 640-960K, and determined that the
more branched isomeric forms of heptane exhibit reduced reactivity. Ignition delay times
of n-butane and iso-butane have been studied in RCM and shock tubes [23-25]. Gersen
et al. [23] measured ignition delay times of n-butane and iso-butane in a RCM in the
temperature range 660K-1010K, and reported that iso-butane exhibits longer ignition
delay times than n-butane below 900 K, and that both isomers results in similar results
above 900 K. Westbrook et al. [26] developed a detailed kinetic model for the heptane
isomers, and it was shown that 2-methylhexane exhibits a notably lower reactivity than n-
heptane in the low and intermediate temperature regimes. Kahandawala et al. [27]
studied ignition delay times in a shock tube for 2-methylheptane. Sarathy et al. [28]
proposed a kinetic model for the oxidation of 2-methylheptane and n-octane, which was
validated against species concentration profiles in a non-premixed 2-methylheptane
counterflow flame, and developed subsequently a comprehensive kinetic model [29] for
the oxidation of 2-methylalkanes up to C
20
.
Systematic studies of isomeric alkanes with 8 to 10 carbon numbers are limited,
especially for lightly branched iso-alkanes, which are known [28] to exist in petroleum-
80
derived jet and diesel fuels. It should be noted that lightly branched, high molecular-
weight iso-alkanes are not readily available and their production cost can be very high.
Additionally, past studies on the ignition of iso-alkanes have been done largely at the
low-temperature regime due to its relevance to knock. On the other hand, flame ignition
that is of relevance to several applications including high-altitude relight in air-breathing
propulsion and which is controlled by high-temperature kinetics, has not been studied
adequately.
Based on these considerations, the goal of this investigation was to study the ignition
of non-premixed flames of C
8
and C
10
isomers and provide further insight into the effects
of the degree and position of branching. The C
8
isomers included 3-methylheptane (3-
MHP), 2,5-dimethylhexane (2,5-DMH), iso-octane, and n-octane, while the C
10
isomers
included 2,7-dimethyloctane (2,7-DMO) and n-decane. Ignition temperatures, T
ign
, were
determined for fuel mole fractions in the fuel/N
2
mixture, X
F
, ranging from 1% to 6%
under atmospheric pressure.
6.2 Experimental Approach
The counterflow approach was used, as described in Chapter 2. Burners with 22 mm
diameter nozzles and 22 mm separation distance were used. The pre-vaporized fuel/N
2
jet was heated to T
u
= 401 K, measured at the fuel-carrying burner exit. For all cases
studied, K = 130 s.
-1
The uncertainty in the reported T
ign
’s was estimated to be ±25 K by considering the
fluctuations of T
air
and the temperature corrections due to radiative and convective heat
81
transfer [30]. In detail, the uncertainty in the reported T
ign
’s resulting primarily from the
temperature correction was between ±6 K and ±9 K [28]. The fluctuations of T
air
and T
u
were estimated to be within ±3 K and ±2 K respectively. The repeatability of the
experimental data was within ±5 K. The injection rates of the liquid fuels were chosen to
be very close to the ignition state, until the uncertainty of X
F
at ignition was within ±2 %.
It was determined that the LDV velocity measurements were determined to be accurate
within ±1%, while the 2σ uncertainty of K to be ±10 s.
-1
Measurements were performed for mixtures of N
2
with n-octane (98%), iso-octane
(99%), 3-MHP (98%), 2,5-DMH (98%), and n-decane (99%), 2,7-DMO (98%)
respectively. Water content of all fuels is less than 0.01%, with the impurities being
largely isomers of similar molecular weight. All octane isomers and decane isomers
considered in the present study accompanying with their boiling temperatures are listed in
table 6.1 and table 6.2.
Table 6.
Chemical n
n-octan
3-methylhe
(3-MHP
2,5-
dimethylhe
(2,5-DM
2,2,4-
trimethylpe
(iso-octan
Table 6.
Chemical n
n-decan
2,7-
dimethyloc
(2,7-DM
.1 Octane
name
ne
ptane
P)
exane
MH)
ntane
ne)
.2 Decane
name
ne
ctane
MO)
e isomers and
Molecular
Formula
C
8
H
18
C
8
H
18
C
8
H
18
C
8
H
18
e isomers an
Molecular
Formula
C
10
H
22
C
10
H
22
d correspond
d correspond
ding boiling t
Structur
ding boiling
Structur
temperature
re
temperature
re
es.
Boiling
Temperat
125 – 126
118 – 120
108 – 110
99 es.
Boiling
Temperat
174
159 – 160
82
g
ture
6
0
0
g
ture
0
6
in
h
fo
co
2
n
st
2
F
nu
th
.3 Mode
The simu
ncluded a hi
ereafter refe
or 2,6-dimet
omponents w
,5-dimethylh
onane, n-de
tudied herein
Figure 6
To illustr
,7-DMO is
igure 6.1.
umber corre
he (2,n-1)-d
eling Appro
ulations for
gh-temperat
erred as Mod
thylheptane
were added
hexane high
ecane, and 2
n could be m
6.1 Carbon
carbon
rate the nam
denoted as C
The carbon
esponding to
dimethylalka
ach
n-octane, 3
ture kinetic m
del I. Mode
and 2,7-DM
to a recently
h temperature
2-methylocta
modeled toge
n skeletal s
n sites labeled
ming of the s
C
10
H
22
-27 in
n chain is l
o the location
anes are sym
3-MHP, 2,5-
model, cons
el I includes
MO. The re
y developed
e oxidation
ane [29] we
ether.
tructure of
d.
species for t
n the model
labeled num
n of the met
mmetrical m
-DMH, 2,7-
sisted of 826
s new high-t
actions for t
d 2-methylhe
model [31,3
ere appende
2,7-dimethy
the (2,n-1)-d
, whose mo
merically (i.e
thyl branch i
molecules.
-DMO and
6 species and
temperature
these (2,n-1
eptane, 3-me
32]. Sub-m
d also, so t
yloctane (C
10
dimethylalka
lecular struc
e., 1,2,3, etc
is the lowest
For 2,7-DM
n-decane fl
d 4725 react
kinetic path
)-dimethyla
ethylheptane
mechanisms f
that all the
0
H
22
-27) wit
anes mechan
cture is show
c.) such tha
t. It is noted
MO species
83
lames
tions,
hways
lkane
e, and
for n-
fuels
th
nism,
wn in
at the
d that
s, the
84
location of a double bond is identified by a hyphen followed by the number of the first
carbon in the double bond, e.g. 2,7-dimethyl-2-octene is C
10
H
20
-2-27. Additional
notations are provided to denote radical sites in the molecule, wherein the carbon sites are
labeled alphabetically (i.e., a,b,c, etc.). In this way, the 2,7-dimethyl-3-octyl radical is
denoted as C
10
H
21
-27c, while the 2,7-methyl-1-octyl radical is written as C
10
H
21
-27a.
The major classes of elementary reactions considered for the oxidation of (2,n-1)-
dimethylalkanes include 10 high temperature reaction classes. These reaction classes and
the selected reaction rate rules for branched alkanes have been described in detail
previously [29]. The THERM [33] software was used to compute the thermochemical
properties of species not present in the 2-methylalkanes model. The THERM group
values are from Benson [34] and Bozzelli and coworkers [35]. The transport data for
(2,n-1)-dimethylalkanes were calculated using the methods documented in [29].
For the simulations of iso-octane, a detailed model developed by Mehl et al. [36] was
adopted as Model II(a), consisting of 874 species and 3796 reactions. The updated core
C
0
-C
4
kinetics of Model I, originally from NUI Galway [37,38], were superimposed on to
Model II(a) and together with the transport data and thermodynamic data of Model I,
Model II(a) was updated and referred to as Model II(b).
Models I and II(b) and the attendant thermodynamic and transport data are available
as supplementary material of Ref. [39].
6
6
K
d
an
fo
b
th
o
ex
M
.4 Resul
.4.1 Igniti
The expe
K = 130 s
-1
a
ecreases as X
nd 2,5-DMH
or iso-octane
etter agreem
he discrepan
ctane/air mi
xperimental
Model II(a) o
lts and Disc
ion of Octan
erimental an
are shown i
X
F
increases
H flames are
e flames com
ment, while u
ncy increasin
ixtures, reve
data (e.g. [
over-predicts
ussion
ne and Deca
nd compute
n Figure 6.
. The comp
e in close agr
mputed using
using Model
ng with X
F
.
aled that M
[31,40]) for
s the data.
ane Isomer F
ed T
ign
’s of
.2. Similar
uted results
reement with
g Model II(a
l II(a) under
Computatio
odel II(b) re
all ’s, esp
Flames
f C
8
isomer
r to n-alkan
using Mode
h the data.
a) and Model
r-predicts the
ons of lamin
esults in bet
pecially for
r flames at
nes [30], T
i
el I T
ign
’s of n
Figure 6.2 d
l II(b). Mod
e data by 13
nar flame sp
tter agreeme
0.9 <
T
u
= 401 K
gn
monoton
n-octane, 3-M
depicts also
del II(b) resu
3 K to 23 K,
peeds,
S
u
o
, o
ent with avai
1.3, while u
85
K and
ously
MHP
T
ign
’s
ults in
, with
f iso-
ilable
using
se
d
co
d
sp
re
M
th
p
Figure 6
Sensitivity
ensitive prim
iffusion to N
onsumption.
iameter and
pecies, such
ecommendat
Model II(b).
he specified
ath of 1-bute
C
4
H
7
1-4 →
C
4
H
7
1-4 →
6.2 Experi
and 2
simula
y analysis o
marily to the
N
2
of iso-but
. The Len
potential we
h as iC
4
H
9
a
tions of Mo
In the core
reverse rate
enyl radical
→ C
4
H
6
+ H
→ C
2
H
4
+ C
imental and
2,5-DMH fl
ation results.
of T
ign
on bin
e diffusion
tene (iC
4
H
8
)
nnard-Jones
ell depth, of
and tC
4
H
9
, w
ourits and R
C
0
-C
4
chem
e of several r
(C
4
H
7
1-4, *
H
C
2
H
3
computed T
lames. Sy
nary diffusio
of fuel to N
) and propen
(L-J) para
f the latter tw
were chang
Rummens [4
mistry of Mo
reactions we
CH
2
CH
2
CH
T
ign
’s of n-oc
ymbols: exp
on coefficie
N
2
, and to l
ne (C
3
H
6
), th
ameters, sp
wo species to
ged by Sarat
1], and thos
odel I, not on
ere updated
H=CH
2
) was
ctane, iso-oct
perimental
ents indicate
lesser extent
he main inte
ecifically th
ogether with
thy et al. [2
se updates w
nly the activ
[29], but als
introduced:
tane, 3-MHP
data. Lines
ed that igniti
t sensitive t
rmediates of
he L-J coll
h a number o
28] based on
were includ
vation energy
so a new rea
(R
(R
86
P,
s:
ion is
to the
f fuel
lision
of key
n the
ded in
y and
action
R1)
R2)
C
M
A
fo
al
pr
ig
a)
=
ag
un
In Model
C
4
H
7
1-4 is c
Model II(a) r
Additionally,
ound that w
ltered, the d
referred to
gnition becau
), which reco
Figure 6.3
= 130 s
-1
alo
greement w
ncertainties.
Figure 6
II(b) 91.5%
consumed vi
esults in low
the rate con
while the bra
ifference in
C-C fission
use 1,3-buta
ombines wit
3 depicts the
ong with the
with the pre
6.3 Experi
Symbo
% of C
4
H
7
1-4
ia R2 to for
wer T
ign
’s an
nstants of R1
anching rati
T
ign
is negli
n, similarly
adiene (C
4
H
6
th H radical a
e experiment
e predictions
sent experim
imental and
ols: experime
4 is consume
rm ethylene
d higher
S
u
o
1 and R2 of W
io between
igible. How
to Miyoshi
6
) results in r
and form no
tal of T
ign
’s o
s using Mod
mental data
computed T
i
ental data. Li
ed via R1, w
(C
2
H
4
) and
o
’s as C
2
H
4
a
Wang et al.
C-H and C
wever, the C-
[43]. On
resonantly st
otably stable
of C
10
isome
del I. The c
a for all X
F
T
ign
’s of n-dec
ines: simulat
while in Mod
d vinyl radi
and C
2
H
3
are
[42] were ad
-C bond sc
-H bond fiss
the other h
tabilized ally
specie C
3
H
6
er flames at T
computed T
F
’s, within t
ane and 2,7-
tion results.
del II(a) 100
cal (C
2
H
3
);
e highly reac
dopted and i
ission is sli
sion herein i
hand, R1 inh
yl radicals (C
6
.
T
u
= 401 K a
T
ign
’s are in
the experim
-DMO flames
87
0% of
thus,
ctive.
it was
ightly
s still
hibits
C
3
H
5
-
and K
close
mental
s.
6
il
k
ag
2
o
h
.4.2 Comp
Figure 6
In Figure
llustrated. I
inetics for C
greements w
,5-DMH, 3-M
f branching
igher T
ign
’s.
parison of Ig
6.4 Experi
octane
Symbo
for n-o
6.4, the effe
n order to p
C
8
isomers,
with the data
MHP, and n
affects the i
Similar tren
gnition Tem
imental and
, ( ■)(---) 3-
ols: experime
octane, 3-MH
ect of the de
provide insig
Model I an
a. The iso-o
n-octane flam
gnition prop
nds can be se
mperatures
computed T
-MHP, (∆)(∙
ental data; L
HP, and 2,5-D
egree of bran
ght into the
nd Model I
octane flame
mes in a desc
pensity, with
een in Figur
T
ign
’s of C
8
is
∙∙∙) 2,5-DMH
Lines: simula
DMH, and M
nching on ig
high-temper
I(b) were u
es exhibit th
cending orde
h higher degr
re 6.5 for the
somers flame
H and ( ◇)( ─
ation results u
Model II(b) fo
gnition of C
8
rature (T >
used as they
he highest T
er, confirmin
rees of branc
e C
10
isomers
es. (○)(-∙-) n
─) iso-octane
using Model
or iso-octane.
isomer flam
1200 K) ign
y result in c
T
ign
’s followe
ng that the d
ching resulti
s flames.
88
n-
e.
I
.
mes is
nition
closer
ed by
egree
ing in
fl
is
si
su
se
co
d
al
Figure 6
Figure 6.6
lames, and th
solate the e
imulations w
ubstituted by
een that 2,7-
onditions th
iffusivity. T
lkyl chains
6.5 Experi
decane
simula
6 depicts the
here is no si
effect of dif
were perform
y those of 2,
-DMO is sli
e increased
Thus, extend
does not h
imental and
e and (○)(---)
ation results u
e experimen
gnificant dif
ffusion and
med in whic
,5-DMH. Th
ightly more
reactivity o
ding of chain
have a mea
computed T
) 2,7-DMO.
using Model
ntal and com
fference in b
d assess the
ch the Lenna
he results ar
reactive tha
f 2,7-DMO
n length by
asurable effe
T
ign
’s of C
10
is
Symbols: e
I.
mputed T
ign
’s
both the data
e effect of
ard-Jones pa
re shown in F
an 2,5-DMH
is counterb
two carbons
fect on T
ign
somers flam
experimental
for 2,5-DM
a and predict
kinetics on
arameters of
Figure 6.6 a
H, so under r
alanced by
s between th
for the X
F
es. ( ◇)( ─) n
l data; Lines
MH and 2,7-D
tions. In ord
n T
ign
, addit
f 2,7-DMO
also, and it c
realistic tran
its slightly l
he two branc
F
-range that
89
n-
s:
DMO
der to
tional
were
an be
nsport
lower
ching
t was
co
si
onsidered in
imilar condit
Figure 6
n this study,
tions [30].
6.6 Experi
(●)(---)
results
change
which is in
imental and
)(∙∙∙) flames.
s using Mode
ed to that of
agreement w
computed T
i
Symbols: e
el I; (∙∙∙) sim
2,5-DMH.
with results
ign
’s of 2,5-DM
experimenta
mulations wit
obtained for
MH ( ◇)( ─) a
al data; Line
th diffusivity
r n-alkanes u
and 2,7-DMO
es: simulatio
y of 2,7-DMO
90
under
O
on
O
Figure 6 6.7 Ranke
compu
d logarithm
uted for a X
F
mic sensitivi
= 4% 3-MH
ity coefficie
HP flame usin
ents of T
ign
ng Model I.
on kinetic
91
cs
Figure 6 6.8 Ranke
compu
d logarithm
uted for a X
F
mic sensitivi
= 4% 2,5-DM
ity coefficie
MH flame us
ents of T
ign
sing Model I.
on kinetic
.
92
cs
Figure 6 6.9 Ranke
compu
d logarithm
uted for a X
F
mic sensitivi
= 4% iso-oct
ity coefficie
tane flame u
ents of T
ign
sing Model I
on kinetic
II(b).
93
cs
C
o
ch
w
Figure 6
The ranke
C
8
isomers fl
ctane flame
hemistry; ho
were identifi
6.10 Rank
comp
ed logarithm
lames at X
F
=
es [30], T
ign
owever, for b
fied. More
ked logarith
puted for a X
mic sensitivit
= 4% and ar
is sensitive
branched iso
eover, in is
hmic sensitiv
X
F
= 4% 2,7-D
ty coefficien
re shown in
e mainly to
omers some
so-octane fl
vity coeffici
DMO flame
nts of T
ign
on
Figures 6.7
H, CO, an
additional s
lames, T
ign
ients of T
ign
using Model
n kinetics we
7, 6.8, and 6
nd C
1
-C
3
sm
mall effects
exhibits so
n
on kinetic
l I.
ere compute
6.9. Similar
mall hydroca
from C
4
kin
ome quantit
94
cs
ed for
to n-
arbon
netics
tative
95
sensitivity to fuel initial consumption reactions as well. Sensitivity analyses show that
the more degree of branching it is the more sensitive of T
ign
to initial fuel consumption
reactions was achieved. The cause was attributed to the lower reactivity of the fuel with
larger degree of branching, to the extent that the reaction rates of several fuel
decomposition paths were low enough to limit the overall reactions preceding ignition, as
evidenced by the reaction rate analyses. In the other word, fuel reactions exhibit
sensitivity when their rates are slow and become thus rate limiting, and this is the case as
the branching degree increases. Sensitivity to fuel reactions was identified also for 2,7-
DMO flames as shown in Figure 6.10.
The highly reactive H and HCO radicals involved in reactions R3 and R4 are key
factors to flame ignition [30].
H + O
2
→ O + OH (R3)
HCO + M → CO + H + M (R4)
It was found, that as the temperature increases, H and HCO concentrations increase
faster in 3-MHP flames, followed by 2,5-DMH, and iso-octane, which is consistent with
the T
ign
trend.
Unlike the ignition of non-premixed n-alkane flames during which large amount of
C
2
H
4
and C
3
H
6
are produced from fuel decomposition via -scission [30], it was
determined that iso-butene (iC
4
H
8
) is a major intermediate during the oxidation of C
8
and
C
10
isomers at the ignition state. The reaction path analyses at the ignition state of 2,5-
DMH and iso-octane flames are shown in Figures 6.11 (a) and (b) respectively.
96
(a)
(b)
Figure 6.11 Reaction path analysis of flames of 2,5-DMH (a) and iso-octane (b)
at the ignition state and X
F
= 4% using (a) Model I, (b) Model II(b).
The numbers indicate the conversion percentages.
fl
fo
co
co
bu
re
F
re
The conce
lames at X
F
=
ound that in
oncentration
onsumed m
utenyl (iC
4
H
esulting in a
igure 6.9. R
eactant and f
iC
4
H
8
+ O
iC
4
H
7
+ H
iC
4
H
7
+ H
Figure 6
entration pro
= 4% at the
ncreasing the
n [31], that
mainly by H,
H
7
) radical vi
an H radical
Reaction R7
forms OH.
OH(O, H) →
H → iC
4
H
8
HO
2
→ iC
4
H
6.12 Conc
octan
ofiles of iC
4
ignition stat
e degree of
t is iso-oc
, O, and OH
ia R5. Subs
l-scavenging
exhibits a p
→ iC
4
H
7
+ H
2
H
7
O + OH
centration pr
ne and 2,7-D
H
8
for 3-MH
te were com
fuel branchi
ctane > 2,5-D
H through
sequently, iC
g loop. The
positive effec
2
O (OH, H
2
)
rofiles of iC
4
DMO flames a
HP, 2,5-DM
mputed and s
ing leads to
DMH ≥ 2,7-D
H abstractio
C
4
H
7
recomb
erefore, R5 i
ct on ignitio
4
H
8
in X
F
= 4
at the ignitio
MH, iso-octan
shown in Fig
o a notable i
DMO > 3-M
on reactions
bines with H
inhibits igni
on as it comp
% 3-MHP, 2
on state.
ne, and 2,7-D
gure 6.12. I
increase in i
MHP. iC
4
H
s and forms
and forms i
ition as show
petes with R
(R
(R
(R
2,5-DMH, iso
97
DMO
t was
iC
4
H
8
H
8
is
s iso-
iC
4
H
8
,
wn in
R6 for
R5)
R6)
R7)
o-
Figure 6
6.13 Conc
MHP
state
centration p
P, 2,5-DMH,
e.
(a)
(b)
rofiles of C
2
, iso-octane
2
H
3
and CH
3
and 2,7-DM
3
in X
F
= 4%
MO flames a
% n-octane, 3
t the ignitio
98
3-
on
99
The concentration profiles of the highly reactive C
2
H
3
and less reactive CH
3
radicals
are shown in Figure 6.13(a) and 6.13(b) respectively. The results are consistent with the
observed ignition propensity of the C
8
and C
10
isomers. n-octane flames produce more
C
2
H
3
compared to the other octane isomers due to its straight chain structure, while the
C
2
H
3
concentrations produced in 2,5-DMH flames are higher compared to iso-octane and
lower compared to 3-MHP. C
2
H
3
reacts primarily via R8 producing vinoxy (CH
2
CHO)
and O radical, which promotes ignition [30] attributing to the quantitative amount of H
radical and H
2
produced subsequently through the paths from CH
2
CHO. Compared to
2,5-DMH, 2,7-DMO flames were found to produce more C
2
H
3
at the ignition kernel.
C
2
H
3
+ O
2
→ CH
2
CHO + O (R8)
Figure 6.13(b) depicts also the concentration profiles of CH
3
radicals at the ignition
state. It can be seen that iso-octane flames produce the largest concentration of CH
3
with
the majority forming from the initial fuel decomposition. In 2,5-DMH flames, CH
3
is
produced mainly from iC
4
H
9
via R9. In 3-MHP flames, sC
4
H
9
and nC
3
H
7
are major
intermediates and CH
3
forms via R10 and R11. In 2,7-DMO flames, nC
3
H
7
forms in
large concentrations and CH
3
is produces mainly via R11 and to lesser extent via R9.
CH
3
is a rather stable radical that can recombine readily via the chain termination
reaction R12 to form the stable molecular C
2
H
6
, thus inhibiting ignition.
iC
4
H
9
→ C
3
H
6
+ CH
3
(R9)
sC
4
H
9
→ C
3
H
6
+ CH
3
(R10)
nC
3
H
7
→ C
2
H
4
+ CH
3
(R11)
2CH
3
+ M → C
2
H
6
+ M (R12)
100
6.5 Concluding Remarks
Ignition temperatures of non-premixed flames of octane and decane isomers were
measured in the counterflow configuration under atmospheric pressure. A detailed
kinetic model was developed for all isomers and the predicted ignition temperatures were
in good agreement with the data.
As expected, it was shown that as the degree of chain branching increases the fuel
reactivity decreases and ignition is inhibited. However, increasing the length of the
straight carbon chain was found to have a non-measurable effect on ignition, as the
influences of fuel diffusivity and reactivity cancel out. Similarly to n-alkane flames, the
ignition temperatures were determined to be sensitive to H
2
/CO and C
1
-C
3
small
hydrocarbons, with the C
4
kinetics starting being relevant as the degree of fuel branching
increases. Analyses of the computed flame structures revealed that the concentrations of
ignition-promoting radicals, such as H, HCO, and C
2
H
3
, and ignition-inhibiting species,
such as CH
3
and iC
4
H
8
, scale consistently with the observed ignition propensities for all
isomers. As the degree of fuel branching increases, ignition-inhibiting intermediates
form directly from fuel-consuming reactions.
6.6 References
[1] C.K. Westbrook, Proc. Combust. Inst. 28 (2000) 1563–1577.
[2] W.G. Lovell, Ind. Eng. Chem. 40 (1948) 2388-2438.
[3] D.J. Vermeer, J.W. Meyer, A.K. Oppenheim, Combust. Flame 18 (1972) 327–336.
101
[4] K. Fieweger, R. Blumenthal, G. Admeit, Proc. Combust. Inst. 25 (1994) 1579–
1585.
[5] R. Minetti, M. Ribaucour, M. Carlier, L.R. Sochet, Combust. Sci. Technol. 113–
114 (1996) 179–192.
[6] R. Minetti, M. Carlier, M. Ribaucour, E. Therssen, L.R. Sochet, Proc. Combust.
Inst. 26 (1996) 747–753.
[7] K. Fieweger, R. Blumenthal, G. Adomeit, Combust. Flame 109 (1997) 599–619.
[8] J.D. Blouch, C.K. Law, Proc. Combust. Inst. 28 (2000) 1679-1686.
[9] D.F. Davidson, M.A. Oehlschlaeger, J.T. Herbon, R.K. Hanson, Proc. Combust.
Inst. 29 (2002) 1295–1301.
[10] S. Tanaka, F. Ayala, J.C. Keck, J.B. Heywood, Combust. Flame 132 (2003) 219–
239.
[11] M.A. Oehlschlaeger, D.F. Davidson, J.T. Herbon, R.K. Hanson, Int. J. Chem.
Kinet. 36 (2) (2004) 67–78.
[12] X. He, M.T. Donovan, B.T. Zigler, T.R. Palmer, S.M. Walton, M.S. Wooldridge,
A. Atreya, Combust. Flame 142 (2005) 266–275.
[13] D.F. Davidson, B.M. Gauthier, R.K. Hanson, Proc. Combust. Inst. 30 (2005)
1175–1182.
[14] X. He, B.T. Zigler, S.M. Walton, M.S. Wooldridge, A. Atreya, Combust. Flame
145 (2006) 552–570.
[15] P.L. Kelly-Zion, J.E. Dec, Proc. Combust. Inst. 28 (2000) 1187–1193.
102
[16] H.J. Curran, P. Gaffuri, W.J. Pitz, C.K. Westbrook, Combust. Flame 129 (2002)
253–280.
[17] D.L. Flowers, S.M. Aceves, J. Martinez-Frias, R.W. Dibble, Proc. Combust. Inst.
29 (2002) 687–693.
[18] S.-C. Kong, R.D. Reitz, Combust. Theory Model. 7 (2) (2003) 417–433.
[19] K.C. Salooja, Combust. Flame 6 (1962) 275–285.
[20] K.C. Salooja, Combust. Flame 9 (1965) 219–227.
[21] M. Ribaucour, R. Minetti, L.R. Sochet, H.J. Curran, W.J. Pitz, C.K. Westbrook,
Proc. Combust. Inst. 28 (2000) 1671–1678.
[22] E.J. Silke, H.J. Curran, J.M. Simmie, Proc. Combust. Inst. 30 (2005) 2639–2647.
[23] S. Gersen, A.V. Mokhov, J.H. Darmeveil, H.B. Levinsky, Combust. Flame 157
(2010) 240–245.
[24] D. Healy, N.S. Donato, C.J. Aul, E.L. Petersen, C.M. Zinner, G. Bourque, H.J.
Curran, Combust. Flame 157 (2010) 1526–1539.
[25] D. Healy, N.S. Donato, C.J. Aul, E.L. Petersen, C.M. Zinner, G. Bourque, H.J.
Curran, Combust. Flame 157 (2010) 1540–1551.
[26] C.K. Westbrook, W.J. Pitz, J.E. Boercker, H.J. Curran, J.F. Griffiths, C. Mohamed,
M. Ribacour, Proc. Combust. Inst. 29 (2002) 1311–1318.
[27] M.S.P. Kahandawala, M.J. DeWitt, E. Corporan, S.S. Sidhu, Energy Fuels 22
(2008) 3673-3679.
[28] S.M. Sarathy, C. Yeung, C.K. Westbrook, W.J. Pitz, M. Mehl, M.J. Thomson,
Combust. Flame 158 (2011) 1277-1287.
103
[29] S.M. Sarathy, C.K. Westbrook, M. Mehl, W.J. Pitz, C. Togbe, P. Dagaut, H.
Wang, M.A. Oehlschlaeger, U. Niemann, K. Seshadri, P.S. Veloo, C. Ji, F.N.
Egolfopoulos, T. Lu, Combust. Flame 158 (2011) 2338-2357.
[30] N. Liu, C. Ji, F.N. Egolfopoulos, Combust. Flame 159 (2012) 465-475.
[31] C. Ji, S.M. Sarathy, P.S. Veloo, C. K. Westbrook, F.N. Egolfopoulos, Combust.
Flame 159 (2012) 1426-1436.
[32] S.M. Sarathy, U. Niemann, C. Yeung, R. Gehmlich, C.K. Westbrook, M. Plomer,
Z. Luo, M. Mehl, W.J. Pitz, K. Seshadri, M.J. Thomson, T. Lu, Proc. Combust.
Inst. 34 (2012), http://dx.doi.org/10.1016/j.proci.2012.05.106.
[33] E. Ritter, J. Bozzelli, Int. J. Chem. Kinet. 23 (1991) 767–778.
[34] S.W. Benson, Thermochemical Kinetics, 2nd ed. Wiley, New york, 1976.
[35] T.H. Lay, J.W. Bozzelli, A.M. Dean, E.R. Ritter, J. Phys. Chem. A 99 (1995)
14514–14527.
[36] M. Mehl, W.J. Pitz, C.K. Westbrook, H.J. Curran, Proc. Combust. Inst. 33 (2011)
193–200.
[37] M.V. Johnson, S.S. Goldsborough, Z. Serinyel, P. O’Toole, E. Larkin, G.
O’Malley, H. J. Curran, Energy Fuels 23 (2009) 5886–5898.
[38] J. Moc, G. Black, J.M. Simmie, H.J. Curran, AIP Conf. Proc. 1148 (2009) 161–
164.
[39] N. Liu, S.M. Sarathy, C.K. Westbrook, F.N. Egolfopoulos, Proc. Combust. Inst.
34 (2012), http://dx.doi.org/10.1016/j.proci.2012.05.040.
104
[40] A.P. Kelley, W. Liu, Y.X. Xin, A.J. Smallbone, C.K. Law, Proc. Combust. Inst.
33 (2011) 501–508.
[41] F.M. Mourits, F.H.A. Rummens, Can. J. Chem. 55 (1977) 3007–3020.
[42] H. Wang, X. You, A.V. Joshi, S.G. Davis, A. Laskin, F.N. Egolfopoulos, C.K.
Law, USC Mech Version II. High-Temperature Combustion Reaction Model of
H2/CO/C1-C4 Compounds. http://ignis.usc.edu/USC_Mech_II.htm, May 2007.
[43] A. Miyoshi, Int. J. Chem. Kinet. 42 (2010) 273-288.
105
Chapter 7
Ignition of Non-Premixed Cyclohexane and Mono-Alkylated
Cyclohexane Flames
7.1 Introduction
Cycloalkanes and alkyl-cycloalkanes are major components of practical fuels.
Recently, more interest has been generated on the chemical kinetics of cyclohexane
(CHX) and mono-alkylated CHX as primary components for diesel fuel surrogates and
jet fuel surrogates.
Several studies have been carried out in jet-stirred reactors (e.g., [1,2]), turbulent flow
reactors (e.g., [3]), shock tubes (e.g., [4-10]), motored engines (e.g., [11]), rapid
compression machines (RCM) (e.g., [12-15]), and flames (e.g., [16]), in order to probe
the low and high temperature kinetics of CHX and methyl-CHX.
Zeppieri et al. [3] studied the pyrolysis of methyl-CHX in a turbulent flow reactor.
Vasu et al. [9] measured OH concentration time-histories during methyl-CHX oxidation
behind reflected shock waves in a heated, high-pressure shock tube. Yang and Boehman
[11] determined the formation of intermediate species and final products during the
106
oxidation of CHX and methyl-CHX at low and intermediate temperature in a motor
engine. Lemaire et al. [12] studied the oxidation and auto-ignition of CHX in a RCM
from 600 K to 900 K, and 0.7 MPa to 1.4 MPa in order to identify the low-temperature
pathways leading to benzene. Davis and Law [16] measured laminar flame speeds for
CHX/air mixtures at ambient temperature and pressure.
Studies on the combustion characteristics of ethyl-CHX, n-propyl-CHX, and
especially for n-butyl-CHX are scant. Ji et al. [17] measured laminar flame speeds of
CHX/air, methyl-CHX/air, ethyl- CHX/air, n-propyl-CHX/air, and n-butyl-CHX/air
mixtures in the counterflow configuration, and showed that CHX/air flames propagate
somewhat faster than the mono-alkylated CHX/air flames largely because of the different
kinetic behavior of the C
3
and C
4
intermediates. Furthermore, no measurable difference
was found among laminar flame speeds of mono-alkylated CHX flames. Vanderover and
Oehlschlaeger [10] measured ignition delay times for methyl-CHX/air and ethyl-CHX/air
mixtures in a shock tube at elevated pressures, and illustrated a relatively higher
reactivity of ethyl-CHX comparing to methyl-CHX due to the addition, as the alkyl group
size increases, of secondary C-H bonds that facilitate H abstraction. Ristori et al. [18]
studied the oxidation of n-propyl-CHX in a jet-stirred reactor at atmospheric pressure.
Crochet et al. [19] investigated the low-temperature oxidation and autoignition of lean n-
propyl-CHX/air mixtures at high pressures (0.45-1.34MPa) in a RCM. Hong et al. [20]
measured ignition delay times behind reflected shock waves for CHX, methyl-CHX, and
n-butyl-CHX at 1.5 and 3 atm, equivalence ratios between 0.5 and 1, and temperatures
between 1280 K and 1480 K. The amount of H radical generated from the decomposition
107
of CHX was found to be higher compared to methyl-CHX, and given that chain-initiation
specific reaction rates for n-butyl-CHX are much faster compared to methyl-CHX,
methyl-CHX reacting mixtures exhibit the largest ignition delay time followed by CHX
and n-butyl-CHX whose ignition behavior is similar. Natelson et al. [21] measured stable
intermediate species during the low-temperature oxidation of n-butyl-CHX in a flow
reactor.
Notably fewer studies exist on flame ignition of CHX and its derivatives. Seshadri
and co-workers studied non-premixed flame ignition of CHX and methyl-CHX in the
counterflow configuration [22-24]. However, there are no flame ignition data for ethyl-
CHX, n-propyl- CHX, and n-butyl- CHX flames.
Several kinetics models have been proposed for CHX and methyl-CHX (e.g., [14, 25-
28]). Dagaut and Cathonnet [29] proposed a kerosene kinetic model including the
chemistry for CHX and n-propyl-CHX. Wang and coworkers [30] developed a kinetic
model for mono-alkylated CHX’s up to n-butyl-CHX, which has been validated against
experimental data for laminar flame speeds, ignition delay times, species profiles behind
reflected shock waves, and species concentration profiles obtained in jet-stirred and flow
reactors.
The goal of this study was to provide archival experimental data for the ignition of
non-premixed flames of CHX, methyl-CHX, ethyl-CHX, n-propyl-CHX, and n-butyl-
CHX, and provide insight into the effect of the fuel structure through detailed numerical
simulations, similarly to a recent study on n-alkanes [31].
108
7.2 Experimental Approach
T
ign
’s were determined in the counterflow configuration at atmospheric pressure and
T
u
= 373 K [32]. The experimental apparatus and procedure are described in detail in
Chapter 2. The separation distance between the two burners and the burner nozzle
diameter was 22 mm each. All experiments were carried out at K = 120 s.
-1
For the
conditions considered in this study, the measured axial velocity gradient at the burner
exits was 60 s.
-1
The flow rates at the burner exits were adjusted to satisfy the
momentum balance and maintain thus the stagnation plane approximately midway
between the two burners. Fuel mole fractions in the fuel/N
2
mixture, X
F
, ranged from 1%
to 10%.
Measurements were performed for mixtures of N
2
with CHX, methyl-CHX, ethyl-
CHX, n-propyl-CHX, and n-butyl-CHX respectively. All fuel reagents were acquired
from TCI America, with water content less than 0.01%. Most of the impurities (<1%)
were isomers of similar molecular weight. All cyclohexane and mono-alkylated
cyclohexane considered in the present study accompanying with their boiling
temperatures are listed in Table 7.1.
Table 7.
Chemica
cycloh
(CH
methyl-cyc
(Methyl
ethyl-cycl
(ethyl-
n-propyl-cy
(n-propy
n-butyl-cyc
(n-butyl
.1 Cycloh
boiling
al name
hexane
HX)
clohexane
l-CHX)
lohexane
CHX)
yclohexane
yl-CHX)
clohexane
l-CHX)
hexane, mon
g temperatur
Molecul
Formul
C
6
H
12
C
7
H
14
C
8
H
16
C
9
H
18
C
10
H
20
no-alkylated
res.
ar
a
0
d cyclohexa
Structur
nes and c
re
correspondin
Boiling
Temperatu
80.74 101 130 – 132
155 181 109
ng
ure
110
7.3 Modeling Approach
T
ign
’s were computed using an opposed-jet flow code and the multi-component
transport coefficient formulation was adopted in all simulations in current chapter.
The recently developed kinetic model JetSurF 2.0 [30], hereafter referred to as Model
I, was used in the simulations. Model I consists of 348 species and 2163 reactions
describing the high-temperature kinetics of all n-alkanes up to n-dodecane, and mono-
alkylated CHX’s up to n-butyl-CHX. For comparison, the model of Dagaut and
Cathonnet [29], consisting of 209 species and 1673 reactions, hereafter referred to as
Model II, was also considered for the CHX flames only.
7
7
se
b
th
X
re
.4 Resul
.4.1 Igniti
Figure 7
Figure 7.1
een that T
ign
oth Model I
he computed
X
F
’s. The hi
ecent study o
lts and Disc
ion of Cyclo
7.1 Experi
data, (─
1 depicts the
n
monotonou
and Model
d ones using
igher reactiv
on flame pro
ussion
ohexane Fla
imental and
─) simulation
e experiment
usly decrease
II are in sat
Model II are
vity of Mode
opagation [17
mes
computed T
ns using Mod
tal and comp
es with X
F
, a
tisfactory ag
e lower than
el II compar
7].
T
ign
’s of CHX
del I, (---) sim
puted T
ign
’s o
as expected
greement wit
n those using
red to Mode
X flames. ( ♦)
mulations usi
of CHX flam
. The comp
th the data f
g Model I esp
el I was con
experimenta
ing Model II
mes, and it c
puted T
ign
’s u
for all X
F
’s, w
pecially at h
nfirmed also
111
al
.
an be
using
while
higher
o in a
112
(a)
(b)
Figure 7.2 Reaction path analysis of a X
F
= 6% CHX flame at the ignition state
using (a) Model I, (b) Model II. The numbers indicate the
conversion percentages.
113
The high-temperature kinetic pathways of CHX were analyzed at the ignition state.
Reaction path analysis was carried out by using Models I and II, and shown in Figures
7.2(a) and 7.2(b) respectively. It can be seen that the same initial fuel consumption
pathways exist in both models. Over 99% of cyclohexane is consumed by H-abstraction
reactions with OH, H, and O to produce cyclohexyl radicals. Most of the cyclohexyl
radicals undergo ring scission reactions and isomerization reactions to form 1-hexen-3-yl
radicals. However, Model I and Model II differ not only in the branching ratios of
unimolecular decomposition of cyclohexyl but also in the production of C
1
-C
3
small
hydrocarbons. In Model II, 98.4% of cyclohexyl radicals undergo initial C-C bond
fission to form 1-hexen-6-yl, and only 1.1% undergoes C-H fission to form cyclohexene
eventually leading to benzene. In Model I, 71% of the cyclohexyl radical forms 1-hexen-
6-yl and 22% produce cyclohexene. During the subsequent β-scissions, C
2
H
4
, C
2
H
3
, and
1,3-butadiene (C
4
H
6
) are produced in Model I, while large amounts of C
2
H
4
and C
2
H
3
are
produced in Model II. The lower ignition propensity of cyclohexane flames predicted by
Model I can be attributed to the stable methyl and allyl radicals produced from 1,3-
butadiene, which is the major intermediate.
Logarithmic sensitivity coefficients (LSC) of T
ign
on reaction rate coefficients
(
i ign
A d T d LSC ln / ln ) for CHX flames are shown in Figure 7.3. Similar to n-alkanes
[31], T
ign
was determined to be sensitive to H
2
/CO and small hydrocarbon kinetics only,
such as:
H + O
2
→ O + OH (R1)
re
it
H + O
2
+
C
4
H
6
+ H
aC
3
H
5
+ H
C
3
H
6
+ H
Figure 7
Note that
eactions. Th
ts reaction r
M → HO
2
+
H + M → SA
H + M → C
3
H → aC
3
H
5
+
7.3 Logari
CHX f
t T
ign
is prim
hat is, R1 ha
rate results i
+ M
XC
4
H
7
+ M
3
H
6
+ M
+ H
2
ithmic sensit
flame, compu
marily sensit
s a positive
in lower T
ig
tivity coeffici
uted using M
tive to the c
effect on T
ig
gn
. Besides,
ients of T
ign
o
Model I.
chain branch
gn
for cycloh
, T
ign
is lim
on kinetics f
hing R1 and
hexane flame
mited also by
(R
(R
(R
(R
for a X
F
= 6%
d terminatio
es, e.g. incre
y the rates o
114
R2)
R3)
R4)
R5)
%
on R2
easing
of H-
co
re
re
C
(a
C
L
m
onsumption
ecombines w
esults in CH
C
2
H
6
. 1,3 P
aC
3
H
5
), whic
Figure 7
LSC’s of
CHX flames
LSC’s of T
ign
magnitude co
reactions
with H via R
H
3
leading to
ropene (C
3
H
ch forms a lo
7.4 Logari
coeffic
T
ign
on bina
are shown in
n
on the CHX
ompared to
involving C
R3 to yield
o the chain
H
6
) produced
oop resulting
ithmic sensi
ients for a X
ary diffusion
n Figure 7.4
X-N
2
BDC a
the attend
C
3
and C
4
1-buten-3-y
termination
d via R4 co
g in H recom
itivity coeffi
X
F
= 6% CHX
n coefficient
. T
ign
is mo
are negative
dant LSC o
4
intermedia
yl (SAXC
4
H
n via the rec
onsumes H
mbination to
ficients of T
X flame, comp
ts (BDC) ( L
st sensitive t
e and approx
n R1. Fur
ates. For
H
7
) radical t
combination
via R5 to f
H
2
.
T
ign
on bina
puted using
ig
T d LSC ln
to the diffus
ximately 3.5
rthermore, b
example,
that subsequ
n of CH
3
to
form allyl ra
ary diffusio
Model I.
m k gn
D d
,
ln /
sion of fuel t
times as lar
being the m
115
C
4
H
6
uently
form
adical
on
m
) for
to N
2
.
rge in
major
in
b
7
n
C
es
b
ntermediate
ecause it con
.4.2 Comp
The ring
-hexane. F
C
6
H
14
flames
specially at
oth CHX an
Figure 7
species of C
nsumes H ra
parison of Ig
structure of
igure 7.5 ill
. It can be
higher X
F
’s
nd n-C
6
H
14
fl
7.5 Experi
Symbo
simula
CHX, the diff
adicals via R
gnition Prop
f cyclohexan
lustrates the
seen that co
. The comp
ames using M
imental and
ols: experim
ation results u
fusion of C
4
H
R3.
pensity of C
ne introduces
e experimen
ompared to n
puted T
ign
’s
Model I.
computed
mental data
using Model
H
6
to N
2
has
Cyclohexane
s specific ch
ntal and com
n-C
6
H
14
, T
ign
are in close
T
ign
’s of CH
( ♦) CHX a
I, (---) CHX
s a negative
e and n-hex
hemical kine
mputed T
ign
n
’s of CHX
agreement
HX and n-C
and ( ▲) n-C
X and ( ─) n-C
effect on ign
xane Flames
etics compar
for CHX an
flames are l
with the dat
C
6
H
14
flames
C
6
H
14
. Lines
C
6
H
14
.
116
nition
s
red to
nd n-
lower
ta for
s.
s:
es
an
tr
d
p
m
w
at
T
ign
is m
specially at
nd 5.742
o
A
ransport coef
iffusivity o
arameters of
modification
while for X
F
<
t higher X
F
’s
Figure 7
most sensitive
low X
F
’s. T
for n-C
6
H
14
fficients are
n T
ign
, sim
f CHX by th
of diffusivi
< 2.5% diffe
s suggesting
7.6 Compu
with o
modifi
e to the fue
The Lennard
4
, indicating
inversely pr
mulations we
hose of n-C
ity of CHX
erences of T
that they are
uted T
ign
of (
riginal Lenn
ed Lennard-
el diffusivit
d-Jones coll
g that CHX
roportional t
ere perform
C
6
H
14
, and th
resulted in
T
ign
between
e caused by
( ─) CHX and
nard-Jones p
-Jones param
ty as discus
lision diame
diffuses fas
to .
2
In ord
med by sub
he results ar
a 15 K inc
the two fuel
kinetics.
d (∙ ‐ ∙) n-C
6
parameters, a
meters of CH
ssed in prev
eters, , are
ster than n-C
der to assess
stituting the
re shown in
crease of T
ig
ls are neglig
6
H
14
flames u
and (---) sim
HX.
vious study
5.29
o
A for
C
6
H
14
, given
the effect o
e Lennard-J
Figure 7.6.
gn
. Addition
gible, they p
using Model
mulations wit
117
[31],
CHX
n that
f fuel
Jones
The
nally,
persist
I
th
Figure 7 7.7 Concen
6%, T
hexane
ntration pro
T
air
= 1223 K
e flames.
(a)
(b)
files of (a) H
K and K = 12
H and HCO,
20s
-1
, (---)cyc
(b) C
2
H
3
and
clohexane fla
d CH
3
at X
F
ames; (—) n
118
=
n-
119
The structure of the ignition kernel was analyzed for CHX and n-C
6
H
14
flames for
X
F
< 6% at the same condition to perceive the size of radical pool prior to ignition. The
concentration profiles of important radicals before ignition are shown in Figures 7.7(a)
and 7.7(b) for X
F
= 6 % at T
air
= 1223 K. Compared to n-C
6
H
14
, H and HCO radicals
produced by the oxidation of CHX are as approximately 3.5 times as those produced by
n-hexane; C
2
H
3
radicals produced by CHX are as 5 times as those produced by n-hexane,
while CH
3
radicals produced by two fuels are similar, which explains the greater ignition
propensity of CHX.
7.4.3 Ignition of Methyl-, Ethyl-, n-Propyl- and n-Butyl-Cyclohexane flames
Figure 7.8 depicts the experimental and computed T
ign
’s for methyl-CHX, ethyl-CHX,
n-propyl-CHX, and n-butyl-CHX flames. In general, the computed T
ign
’s using Model I
are systematically slightly higher compared to the data but within experimental
uncertainty, except for n-butyl-CHX for which the data are over-predicted by about 45 K.
The experimental T
ign
’s of n-C
6
H
14
, methyl-CHX, ethyl-CHX, n-propyl-CHX, and n-
butyl-CHX flames are shown in Figure 7.9. T
ign
’s of all mono-alkylated CHX flames are
very close to each other and similar to n-C
6
H
14
, while T
ign
’s of CHX flames are lower.
The simulations capture also the experimentally observed trend. These results
demonstrate that the presence of the alkyl group on the ring structure, reduces the ignition
propensity but the size of alkyl group has no measurable effect, in agreement with
observations made for flame propagation [17].
120
Reaction path analyses revealed that CHX, methyl-CHX, and mono-alkylated CHX
with side-chain length larger than two, follow different fuel consumption pathways at the
ignition state. CHX undergoes H-abstraction and forms cyclohexyl that undergoes
largely intermolecular isomerization to form its straight chain isomers. Unlike the
reaction pathways in premixed methyl-CHX/air flames [17] in which 100% methyl-CHX
is consumed by H-abstraction to form five types of methyl-cyclohexyl radicals, at the
ignition state 41% of methyl-CHX undergoes direct intermolecular isomerization to
produce its straight chain isomers, and 59% undergoes H-abstraction to form various
methyl-cyclohexyl radicals. Due to the stability of the CH
3
group, only 0.1% of the fuel
undergoes scission on the CH
3
chain to form methyl and cyclohexyl radicals. The
pathways of ethyl-CHX, n-propyl-CHX, and n-butyl-CHX were determined to be similar.
Figure 7.10 depicts the reaction pathway of an ethyl-CHX flame at the ignition state.
74% of ethyl-CHX undergoes ring scission under which 37% directly forms its straight
chain isomers and 37% forms the branch chain isomers. 19% of ethyl-CHX undergoes
scission of the alkyl group to form ethyl and cyclohexyl radicals. Subsequently,
cyclohexyl radicals follow the same route as in cyclohexane oxidation. Large amounts of
C
1
-C
4
small hydrocarbons are produced by the straight chain isomers of ethyl-CHX. The
reaction rate for ethyl-CHX decomposition to produce the straight chain isomer is about
250 times greater compared to that resulting in branch chain isomers. Similar pathways
were determined for n-propyl-CHX and n-butyl-CHX at the ignition state.
Figure 7 7.8 Experi
propyl
data. L
imental and
l-CHX, and
Lines: simula
computed T
n-butyl-CH
ation results u
T
ign
’s for met
HX flames.
using Model
thyl-CHX, e
Symbols:
l I.
ethyl-CHX, n
experimenta
121
n-
al
Figure 7
Figure 7
7.9 Compa
CHX,
butyl-C
7.10 Reac
ignit
perc
arison of ex
( □) methyl-
CHX, and ( ▲
ction path a
tion state usi
entages.
perimentally
-CHX, ( ∆) e
▲) n-C
6
H
14
.
analysis of a
ng Model I.
y determined
ethyl-CHX, (
a X
F
= 6%
The number
d T
ign
’s for
( ○) n-propyl
ethyl-CHX
rs indicate th
flames of ( ♦
l-CHX, (
*
) n
flame at th
he conversio
122
♦)
n-
he
on
an
sh
et
m
st
re
C
Reaction
nd as repres
hown in Fig
thyl-CHX,
methyl-CHX
tructure redu
eactivity is in
CHX, in agre
Figure 7
rates of initi
entatives tho
ure 7.11. It
n-propyl-CH
that is the
uces the fue
ndependent
eement with
7.11 Reac
prem
ignit
CHX
ial C-C bond
ose reaction
was found t
HX, and n-
least reacti
l reactivity,
of the chain
the findings
ction rates
mixed CHX
tion state for
X; and (—) n
d fission at t
rates for CH
that CHX ex
-butyl-CHX
ive. Thus,
while for si
length but l
of Vandero
of initial C
, methyl-CH
X
F
= 6% an
-propyl-CHX
the ignition
HX, methyl-C
xhibits the h
that exhib
the methyl
ide-chain len
lower than C
ver and Oeh
C-C bond fi
HX and n
nd K = 120s
-1
,
X.
state were c
CHX, and n
highest reacti
bit similar
l group add
ngth larger t
CHX and gre
hlschlaeger [
ission compu
-propyl-CHX
, (‒∙∙‒) CHX
computed as
-propyl-CHX
ivity followe
reactivities,
dition to the
than two the
eater than me
[10].
uted of non
X flames a
; (---) methy
123
s well
X are
ed by
, and
e ring
e fuel
ethyl-
n-
at
l-
al
re
sm
sp
Sensitivity
lkylated CH
espectively.
mall hydroca
pecific to tho
Figure 7
y analyses
HX flames at
Similar to
arbon kineti
ose alkylated
7.12 Loga
mon
of T
ign
on
t X
F
= 6% a
CHX flames
cs but not to
d fuels.
arithmic sens
o-alkylated C
kinetics an
and the resu
s, T
ign
is sen
o reactions in
sitivity coeff
CHX flames,
nd BDC’s
ults are show
nsitive to H
2
nvolving the
ficients of T
ig
, computed u
were perfor
wn in Figur
, CO, C
2
H
3
,
e fuel and lar
gn
on kinetics
using Model
rmed for m
es 7.12 and
, CH
3
, and a
rge intermed
s for X
F
= 6%
I.
124
mono-
d 7.13
aC
3
H
5
diates
%
fu
C
6
b
d
sl
re
C
th
Figure 7
The effec
uel-N
2
BDC
CHX, methyl
.00
o
A, 6.03
inary diffus
iffusivities o
lightly highe
eactivity of
CHX’s. Thu
heir similar
7.13 Loga
coeff
using
ct of diffusio
’s having the
l-CHX, ethy
2
o
A, and 6
sivity. Thus
of ethyl-CHX
er than n-bu
CHX result
s, T
ign
’s of e
diffusivity a
arithmic sen
ficients for X
g Model I.
on on T
ign
is
e largest pro
yl-CHX, n-p
.229
o
A
respe
s, CHX diff
X and n-pro
utyl-CHX.
t in higher i
ethyl-CHX a
and reactivit
nsitivity coef
X
F
= 6% mo
very similar
omoting effec
propyl-CHX,
ectively
[30]
ffuses faster
opyl-CHX ar
The highest
ignition prop
and n-propyl
ty. On the o
fficients of
ono-alkylated
r for all mon
ct on ignitio
, and n-buty
], which is
than all m
re similar, l
t diffusivity
pensity com
l-CHX flame
other hand,
T
ign
on bin
d CHX flam
no-alkylated
on. The Lenn
yl-CHX are
inversely pr
mono-alkylate
ower than m
y combined
mpared to all
es are nearly
the similar
ary diffusio
mes, compute
d CHX’s wit
nard-Jones
5.29
o
A, 5.6
roportional t
ed CHX’s.
methyl-CHX
with the hi
l mono-alky
y identical d
T
ign
’s of me
125
on
ed
th the
’s of
68
o
A ,
to the
The
X, and
ighest
ylated
due to
ethyl-
126
CHX flames compared to ethyl-CHX and n-propyl-CHX is a result of the combined
effects of lower reactivity and higher fuel diffusivity on ignition. This was confirmed by
ignition kernel analyses, which revealed that the concentrations of key radicals in methyl-
CHX flames are very similar to those for ethyl-CHX and n-propyl-CHX flames. Finally,
the slightly higher computed T
ign
’s for n-butyl-CHX flames could be attributed to the
synergistic effects of lower concentrations of reactive small species, such as C
2
H
3
, and
the slightly lower fuel diffusivity compared to ethyl-CHX and n-propyl-CHX. C
2
H
3
results directly from C
2
H
4
through H-abstraction, with C
2
H
4
being the main product of n-
C
3
H
7
decomposition in both
n-propyl-CHX and n-butyl-CHX flames. The consumption
of n-propyl-CHX and n-butyl-CHX results in a number of straight chain C
9
and C
10
isomers. It was determined that compared to the C
10
, the decomposition of the C
9
isomers results in higher concentrations of n-C
3
H
7
and thus C
2
H
4
and C
2
H
3
. Additionally,
C
2
H
4
is produced to a lesser extent from the direct scission of the alkylated group from
the ring structure. It was found that this secondary pathway results in more C
2
H
4
in n-
propyl-CHX flames.
The concentration profiles of the highly active H, HCO, C
2
H
3
and low reactive CH
3
radicals are shown in Figure 7.14. The results do provide concrete evidence regarding
the cause of the higher ignition propensity for CHX flames and similar ignition
propensity for methyl-CHX, ethyl-CHX, and n-propyl-CHX flames. H, HCO and C
2
H
3
produced from methyl-CHX is similar as those produced from ethyl- and n-propyl-CHX
due to the competition between the fuel reactivity and diffusivity, while those produced
fr
m
rom n-butyl
mono-alkylat
-CHX are t
ted CHXs.
the least tha at explained
(a)
(b)
d the higher r computed T
ign
’s than
127
other
Figure 7 7.14 Conc
X
F
=
alkyl
centration pr
6%, T
air
=
lated cyclohe
(c)
(d)
rofiles of (a)
1223 K and
exane flames
H, (b) HCO
d K = 120s
-1
s.
O, (c) C
2
H
3
, a
for non-pre
nd (d) CH
3
a
emixed mono
128
at
o-
129
7.5 Concluding Remarks
Ignition temperatures of non-premixed cyclohexane and mono-alkylated cyclohexane
flames were measured in the counterflow configuration under atmospheric pressure, a
constant strain rate of 120 s,
-1
and at an unburned fuel/N
2
mixture temperature of 373 K.
The flow velocities were determined using laser Doppler velocimetry and the reported
strain rate was measured locally on the fuel side. The recently developed JetSurF 2.0
detailed kinetic model consisting of 348 species and 2163 reactions was used in the
numerical simulations for all flames. The computed ignition temperatures are in close
agreement with the data for cyclohexane, methyl-, ethyl-, and n-propyl-cyclohexane
flames, but they over-predict the n-butyl-cyclohexane data by about 20 K.
It was determined that cyclohexane flames exhibit the highest ignition propensity
among all mono-alkylated cyclohexanes due to its higher reactivity and diffusivity. On
the other hand, the size of alkyl group chain was found experimentally to have no
measurable effect on ignition, and sensitivity and reaction path analyses showed that this
behavior stems from the competition between fuel reactivity and diffusivity. However,
while using the JetSurF 2.0 model reproduces in general this trend, a reduced ignition
propensity of n-butyl-cyclohexane flames is predicted, due to the synergistic effects of
lower concentrations of reactive small species, such as C
2
H
3
, and the slightly lower fuel
diffusivity compared to the other mono-alkylated cyclohexanes. Similarly to n-alkane
flames, the ignition temperatures are sensitive to H
2
/CO and C
1
-C
3
small hydrocarbons.
130
7.6 References
[1] D. Voisin, A. Reuillon, J.-C. Boettner, Combust. Sci. Technol. 138 (1998) 137–
158.
[2] A.El Bakali, M. Braun-Unkhoff, P. Dagaut, P. Reank, M. Cathonnet, Proc.
Combust. Inst. 28 (2000) 1631–1638.
[3] S. Zeppieri, K. Brezinsky, I. Glassman, Combust. Flame 108 (1997) 266–286.
[4] A. Burcat, R.C. Farmer, R.L. Espinoza, R.A. Matula, Combust. Flame 36 (1979)
313–316.
[5] J.P. Orme, H.J. Curran, J.M. Simmie, J. Phys. Chem. A 110 (2006) 114–131.
[6] B. Sirjean, F. Buda, H. Hakka, P.A. Glaude, R. Fournet, V. Warth, F. Battin-
Leclerc, M. Ruiz-Lopez, Proc. Combust. Inst. 31 (2007) 277–284.
[7] S.E. Daley, A.M. Berkowitz, M.A. Oehlschlaeger, Int. J. Chem. Kin. 40 (2008)
624–634.
[8] S.S. Vasu, D.F. Davidson, Z. Hong, R.K. Hanson, Energy Fuels 23 (2009) 175–
185.
[9] S.S. Vasu, D.F. Davidson, R.K. Hanson, Combust. Flame 156 (2009) 736–749.
[10] J. Vanderover, M. Oehlschlaeger, Int. J. Chem. Kin. 41 (2009) 82–91.
[11] Y. Yang, A.L. Boehman, Proc. Combust. Inst. 32 (2009) 419–426.
[12] O. Lemaire, M. Ribaucour, M. Carlier, R. Minetti, Combust. Flame 127 (2001)
1971–1980.
[13] S. Tanaka, F. Ayala, J.C. Keck, J.B. Heywood, Combust. Flame 132 (2003) 219–
239.
131
[14] W.J. Pitz, C.V. Naik, T. Ní Mhaoldúin, C.K. Westbrook, H.J. Curran, J.P. Orme,
J.M. Simmie, Proc. Combust. Inst. 31 (2007) 267–275.
[15] G. Mittal, C.-J. Sung, Combust. Flame 156 (2009) 1852–1855.
[16] S.G. Davis, C.K. Law, Combust. Sci. Technol. 140 (1998) 427–449.
[17] C. Ji, E. Dames, B. Sirjean, H. Wang, F.N. Egolfopoulos, Proc. Combust. Inst. 33
(2011) 971–978.
[18] A. Ristori, P. Dagaut, A. El Bakali, M. Cathonnet, Combust. Sci. Technol. 165
(2001) 197–228.
[19] M. Crochet, R. Minetti, M. Ribaucour, G. Vanhove, Combust. Flame 157 (2010)
2078-2085.
[20] Z. Hong, K.-Y. Lam, D.F. Davidson, R.K. Hanson, Combust. Flame 158 (2011)
1456-1468.
[21] R.H. Natelson, M.S. Kurman, N.P. Cernansky, D.L. Miller, Combust. Flame 158
(2011) 2325-2337.
[22] S. Humer, A. Frassoldati, S. Granata, T. Faravelli, E. Ranzi, R. Seiser, K. Seshadri,
Proc. Combust. Inst. 31 (2007) 393–400.
[23] K. Seshadri, S. Humer, R. Seiser, Combust. Theor. Model. 12 (2008) 831–855.
[24] T. Bieleveld, A. Frassoldati, A. Cuoci, T. Faravelli, E. Ranzi, U. Niemann, K.
Seshadri, Proc. Combust. Inst. 32 (2009) 493–500.
[25] S. Granata, T. Faravelli, E. Ranzi, Combust. Flame 132 (2003) 533–544.
[26] F. Buda, B. Heyberger, R. Fournet, P.-A. Glaude, V. Warth, F. Battin-Leclerc,
Energy Fuels 20 (2006) 1450-1459.
132
[27] C. Cavallotti, R. Rota, T. Faravelli, E. Ranzi, Proc. Combust. Inst. 31 (2007) 201–
209.
[28] E.J. Silke, W.J. Pitz, C.K. Westbrook, M. Ribaucour, J. Phys. Chem. A 111 (2007)
3761–3775.
[29] P. Dagaut, M. Cathonnet, Prog. Energy Combust. Sci. 32 (2006) 48–92.
[30] H. Wang, E. Dames, B. Sirjean, D. A. Sheen, R. Tangko, A. Violi, J. Y. W. Lai, F.
N. Egolfopoulos, D. F. Davidson, R. K. Hanson, C. T. Bowman, C. K. Law, W.
Tsang, N. P. Cernansky, D. L. Miller, R. P. Lindstedt, A high-temperature
chemical kinetic model of n-alkane (up to n-dodecane), cyclohexane, and methyl-,
ethyl-, n-propyl and n-butyl-cyclohexane oxidation at high temperatures, JetSurF
version 2.0, September 19, 2010 (http://melchior.usc.edu/JetSurF/JetSurF2.0).
[31] N. Liu, C. Ji, F.N. Egolfopoulos, Combust. Flame 159 (2012) 465-475.
[32] N. Liu, C. Ji, F.N. Egolfopoulos, Proc. Combust. Inst. 34 (2012) doi:
10.1016/j.proci.2012.05.104.
133
Chapter 8
Ignition Characteristics of Non-premixed Binary Fuel Blends Flames
8.1 Introduction
Among the limiting factors in the operation of hydrocarbon-fueled scramjets is the
ignition process in a high-speed environment. High-density liquid hydrocarbons are
expected to serve also as coolants for the fuselage and various engine components. Thus,
the fuel could be heated up at high temperatures and eventually undergo thermal cracking
at various extents (e.g., [1-3]). Ethylene (C
2
H
4
) is among the major products of cracking
of the kerosene-type fuels, among which n-dodecane (n-C
12
H
26
) is a representative
constituent.
The computed major products of n-C
12
H
26
decomposition are identified mainly as H
2
,
CH
4
, and C
2
H
4
under temperatures of 500 K, 800 K, and 1000 K are shown in Figure 8.1.
At 1000 K, C
2
H
4
accounts for approximately 25 % of the decomposition products.
134
Figure 8.1 Major products of n-dodecane decomposition under 500 K, 800 K,
and 1000 K.
Relatively few studies on fuel blends exist on flame propagation [4-12] and flame
ignition [11-16] with emphasis on H
2
, CH
4
, or C
2
H
4
.
Egolfopoulos and Dimotakis [1-3] investigated numerically the ignition and laminar
flame speeds of C
2
H
4
and its blends with CH
4
, H
2
, F
2
, and NO. Hirasawa et al. [7]
studied laminar flame speeds of mixtures of air with C
2
H
4
, n-butane, toluene, and their
binary blends using the counterflow twin flames technique and developed a semi-
empirical mixing rule based on the results that the flame temperature has the dominant
influence on the flame speeds of the fuel blends and minor effect of chemical kinetic
coupling was realized. The promoting effects of H
2
and C
2
H
4
on CH
4
flame ignition have
135
been documented in counterflow flame studies (e.g. [11,13]). Aggarwal et al. [16]
carried out a numerical investigation on the ignition of mixtures H
2
or CH
4
with n-C
7
H
16
,
in a homogenous reactor. The results show that H
2
or CH
4
addition has a relatively small
effect on the ignition of n-C
7
H
16
/air mixtures, while the ignition of H
2
/air mixtures and
CH
4
/air mixtures is influenced significantly by even small amount of n-C
7
H
16
addition,
indicating that the ignition behaves of those fuel blends are controlled by the oxidation of
n-C
7
H
16
.
Extensive kinetic models have been developed for the oxidation of C
2
H
4
and several
mechanisms have been proposed for n-C
12
H
26
. Marinov et al. [17] developed a model for
aromatics and PAH formation in a fuel-rich, n-C
4
H
10
/oxygen/argon, atmospheric flame
and a fuel-rich C
3
H
8
flame, which has been validated against data for fuel-rich C
2
H
6
,
C
2
H
4
, and CH
4
flames. Wang et al. [18] proposed a kinetic model for high-temperature
oxidation of H
2
, CO, and C
1
-C
4
hydrocarbons, which has been tested against a wide range
of combustion data. Petersen et al. [19] developed a kinetic mechanism for natural gas,
and Bourque et al. [20] updated the model subsequently. Ranzi et al. [21] developed a
lumped model for investigating the oxidation of heavy n-alkanes from n-C
10
H
22
to n-
C
16
H
34
. You et al. [22] proposed a high-temperature detailed kinetic model for the
combustion of n-alkanes up to n-C
12
H
26
and has been validated against experimental data
for laminar flame speeds, ignition delay times, species profiles behind reflected shock
waves, and flow reactors. Recently, Westbrook et al. [23] developed a detailed kinetic
model describing the oxidation of n-alkanes from n-C
8
H
18
to n-C
16
H
34
.
fl
pr
C
1
8
co
pr
n
th
Regarding
lames of hig
resent study
C
2
H
4
, n-C
12
H
: 9, 1 : 4, an
.2 Exper
Similarly
onfiguration
Figure 8
Briefly, a
re-vaporized
ozzle diame
he pre-vapor
g the study
h molecular
y aims to pro
H
26
, and C
2
H
nd 1 : 1 on a
rimental Ap
to previous
n, as schemat
8.2 Schem
a hot air strea
d n-C
12
H
26
/C
eters and the
rized n-C
12
H
of fuel blen
weight hydr
ovide archiva
H
4
/n-C
12
H
26
b
volume basi
pproach
s chapters, t
tically shown
matic of the ex
am is directe
C
2
H
4
/N
2
mix
burner sepa
H
26
/N
2
mixtu
nds, the effe
rocarbons ha
al experimen
blends, with
is.
the experime
n in Figure 8
xperimental
ed downwar
tures exiting
aration distan
ure and its f
ects of C
2
H
4
ave not been
ntal data for
blending C
2
ents were p
8.2.
configuratio
rd from the u
g from the lo
ance were 22
flow rate wa
4
addition o
n studied sys
the ignition
2
H
4
: n-C
12
H
erformed in
on.
upper quartz
ower burner.
2 mm. C
2
H
4
as controlled
on the ignitio
stematically.
of non-prem
H
26
ratios equ
n the counter
z burner aga
. Both the b
4
was mixed
d by a mass
136
on of
. The
mixed
ual to
rflow
inst a
burner
d with
flow
137
meter or using sonic nozzles with various sizes by controlling the upstream pressure of
the orifice according to the flow rate ranges requiring in the experiment. Under all
experimental conditions, T
u
= 453 K and K = 130 s.
-1
Ignition was achieved by establishing the appropriate flow rates of the pre-vaporized
n-C
12
H
26
/C
2
H
4
/N
2
mixtures and air at the two burner exits and by gradually increasing
T
air
until a flame appeared. The reported T
ign
include correction for radiation/convective
heat losses.
Measurements were performed for mixtures of N
2
diluted n-C
12
H
26
(99%) and C
2
H
4
(99%). n-C
12
H
26
was acquired from TCI America, with water content less than 0.01%.
Most of the impurities were isomers of similar molecular weight.
8.3 Modeling Approach
In all simulations, the stretch-corrected T
ign
’s were obtained by changing the value of
the velocity gradient, , at the burner exit to conform with the experimental strain rate
distribution. Note that stretch corrections can result in the reduction of the computed T
ign
by as much as 20 K.
Three kinetic models were used in the present study to simulate the ignition of non-
premixed C
2
H
4
flames, including USC Mech II consisting of 111 species and 784
reactions, developed by Wang et al. [18], a kinetic mechanism for natural gas referred as
NUI Mech consisting of 132 species and 821 reactions, reported by Bourque et al. [20],
and a Complete San Diego Mechanism (UCSD Mech) released by 2011 containing 50
138
species and 244 reactions [24]. JetSurF 1.0 [25], which is based on USC Mech II, was
adopted in the simulations of n-C
12
H
26
and n-C
12
H
26
/C
2
H
4
non-premixed flames.
8.4 Results and Discussion
8.4.1 Ignition of Non-Premixed Flames of C
2
H
4
and Validation of Detailed Kinetic
Models
Figure 8.3 illustrates the experimental and computed T
ign
’s for C
2
H
4
flames. It can be
seen that T
ign
’s predicted using NUI Mech and UCSD Mech are in generally close
agreement with the experimental data within the accuracy of the measurements. USC
Mech II predicts consistently higher T
ign
’s than NUI Mech and UCSD Mech.
Additionally; the overprediction of T
ign
by USC Mech is approximately 50 K to 70 K and
becomes larger as X
F
decreases. Liu et al. [11] also reported an about 50 K over-
prediction by USC Mech II for non-premixed C
2
H
4
flames at K = 300 s,
-1
p = 1atm,
T
u
= 298 K, and C
2
H
4
concentration of 20% by volume, which is consistant with the
measurements in the present study.
The logarithmic sensitivity coefficients (LSC’s) of T
ign
’s to reaction-rate constants for
a X
F
= 6% N
2
diluted C
2
H
4
flames computed using USC Mech II, UCSD Mech and NUI
Mech are shown in Figure 8.4. The analysis indicates that the dominant reactions
impacting T
ign
for USC Mech II are the C
2
H
4
abstraction reaction R1, the vinyl radical
(C
2
H
3
) oxidation reactions R2 and R3, the heat release reactions and chain termination
reactions involving H/OH/CO. On the other hand, T
ign
is primarily sensitive to the chain
branching reaction R4 in NUI Mech instead of R1 in USC Mech II as seen in Figure 8.4.
139
C
2
H
4
+ OH → C
2
H
3
+ H
2
O (R1)
C
2
H
3
+ O
2
→ CH
2
CHO + O (R2)
C
2
H
3
+ O
2
→ CH
2
O + HCO (R3)
H + O
2
→ O + OH (R4)
It is also if interest to note that T
ign
exhibits large sensitivity on reaction R5 that is
included in NUI Mech and UCSD Mech but not in USC Mech II.
C
2
H
4
+ O → CH
2
CHO + H (R5)
C
2
H
4
+ O → CH
3
+ HCO (R6)
R5 promotes ignition because it produces highly reactive H radical and vinyoxy
radicals (CH
2
CHO). Egolfopoulos and Dimotakis [1] reported that omitting CH
2
CHO
and its reactions results in the underestimation of the ignition propensity of C
2
H
4
flames.
R6 is the primary channel for C
2
H
4
reacting with O radical, and it competes reactants
with R5 in NUI Mech and UCSD Mech thus inhibiting ignition. When adding R5 to
USC Mech II, the computed T
ign
using the rate parameters of R5 from NUI Mech and
UCSD Mech were similar, and were determined to be lower by approximately 12 K for
X
F
= 6% N
2
-diluted C
2
H
4
flames comparing to that using USC Mech II. Furthermore, the
addition of R5 to USC Mech II was determined to have no measurable influences on the
prediction of flame propagation. Therefore, the inclusion of both R5 and R6 in the
mechanism was considered to achieve better prediction on T
ign
applied herein, whereas it
does not affect the prediction of laminar flame speeds of C
2
H
4
. Similar observations
were made by Varatharajan and Williams [26] as well for properly lowering predicted
ig
2
gnition time
500 K.
Figure 8
s of C
2
H
4
in
8.3 Experi
n a shock tu
imental and
ube covered
computed T
i
d initial tem
ign
’s of non-p
mperatures be
premixed C
2
H
etween 1000
H
4
flames.
140
0 and
st
d
re
w
Figure 8
Reaction
tate using U
ecompositio
eacting with
was noted tha
8.4 Logari
coeffic
X
F
= 6%
path analys
USC Mech II
on, the main
O radical an
at R5 and R
ithmic sens
ients compu
%.
ses were per
I, UCSD Me
differences
nd CH
2
CHO
R7 are maint
sitivity coef
uted for C
2
rformed for
ech, and NU
in the predi
O oxidation k
tained in UC
fficients of
2
H
4
flames;
C
2
H
4
flame
UI Mech. T
ictions are th
kinetics amo
CSD Mech b
T
ign
on r
T
u
= 453 K
es at X
F
= 6%
hroughout th
he reaction c
ong the three
but not in U
reaction rat
K, K = 130 s
-
% at the ign
he paths of
channels of
e mechanism
USC Mech II
141
te
-1
,
nition
C
2
H
4
C
2
H
4
ms. It
I. R5
an
o
is
U
du
nd R7 are th
f H produce
s R5.
CH
2
CHO
The main
UCSD Mech,
CH
2
CHO
CH
2
CHO
CH
2
CHO
Figure 8
Comparin
uring the ox
he main rout
d from those
→ CH
2
CO
n paths of CH
, and NUI M
→ CH
2
CO
→ CH
2
CO
→ CH
2
O →
8.5 CH
2
O
compu
ng to USC M
xidation of C
tes of H rad
e two reactio
+ H
H
2
CHO oxi
Mech and sho
→ HCCO →
→ CH
2
OH
→ HCO → C
concentrat
uted for C
2
H
4
Mech II, for
CH
2
CHO in
dical product
ons, and in N
dation at ign
own as P1, P
→ CO → CO
→ CH
2
O →
CO → CO
2
ion profiles
4
flames at T
a
rmaldehyde
UCSD Mec
tion in UCSD
NUI Mech th
nition were
P2, and P3 re
O
2
→ HCO → C
s along the
T
air
= 1173 K a
(CH
2
O) is a
ch and NUI
D Mech, sp
he primary H
identified fo
espectively.
O → CO
2
e stagnation
and X
F
= 6%
a major inte
I Mech, whi
ecifically 52
H production
(R
for USC Mec
(P
(P
(P
n streamlin
%.
ermediate sp
ch underwen
142
2.4 %
n path
R7)
ch II,
P1)
P2)
P3)
ne
pecies
nt H-
ab
co
pr
b
d
bstraction r
oncentration
roduction of
CH
2
O + H
Figure 8
The value
inary diffusi
iluted C
2
H
4
eaction resu
n profiles of
f amount of H
H → HCO +
8.6 Logari
coeffic
X
F
= 6%
es of the co
ion coefficie
flames is m
ulting in th
f CH
2
O form
H and H
2
fac
+ H
2
ithmic sensi
ients compu
%.
omputed log
ents (BDC’s)
mainly sensi
he primary
med were c
cilitates igni
itivity coeffi
uted for C
2
garithmic sen
) as shown i
itive to the
production
computed as
ition.
ficients of T
2
H
4
flames;
nsitivity coe
in Figure 8.6
diffusion o
path for H
s shown in F
T
ign
on bina
T
u
= 453 K
efficients (L
6, suggest th
f C
2
H
4
-N
2
p
H
2
via R8.
Figure 8.5.
(R
ary diffusio
K, K = 130 s
-
LSC’s) of T
i
hat ignition o
pair as expe
143
The
The
R8)
on
-1
,
ign
on
of N
2
-
ected.
144
Additionally, LSC’s of T
ign
on the C
2
H
4
-N
2
BCD are about twice as large in magnitude
compared to kinetics, emphasizing the importance of transport in non-premixed flame
ignition. Comparing the transport properties in the three mechanisms, the Lennard-Jones
collision diameter of C
2
H
4
in USC Mech II is 0.475 Å larger than those in UCSD Mech
and NUI Mech in which C
2
H
4
diffusivities are the same. By changing the transport
parameters of C
2
H
4
in USC Mech II to that in NUI Mech, T
ign
decreases by 17 K, similar
as Liu et al. [11] reported, which along with the inclusion of R5 contributes large
proportion to the difference in the simulations using the three models.
8.4.2 Ignition of Non-Premixed Flames of n-C
12
H
26
, and n-C
12
H
26
/C
2
H
4
Blends
Figure 8.7 depicts the measured and computed using JetSurF 1.0 T
ign
’s for flames of
n-C
12
H
26
and n-C
12
H
26
/C
2
H
4
with 1:9, 1:4, and 1:1 blending ratios. The simulation
results for n-C
12
H
26
flames are in good agreement with the experimental data. It is of
interesting to note that the discrepancy becomes larger with the increasing C
2
H
4
concentration in the fuel blend, and such discrepancy originates from the over-prediction
of T
ign
’s of C
2
H
4
flames under all X
F
’s addressing the need of further investigation on the
chemistry of high-temperature ignition of C
2
H
4
flames.
8
b
as
Figure 8
.4.3 Effect
The exper
lends flames
s expected b
8.7 Experi
blends
results
ts of Ethyle
rimentally d
s are compar
between thos
imental and
flames. S
s.
ne on the Ig
etermined T
red in Figur
se of C
2
H
4
an
computed T
Symbols: ex
gnition of n-
T
ign
’s of flam
e 8.8 as func
nd n-C
12
H
26
T
ign
’s of n-C
xperimental
-Dodecane F
mes of C
2
H
4
,
ctions of X
F
flames.
C
12
H
26
, and n
data. Lines
Flames
n-C
12
H
26
, an
. The data o
n-C
12
H
26
/C
2
H
s: simulatio
nd n-C
12
H
26
/
of fuel blend
145
H
4
on
/C
2
H
4
ds are
b
an
ad
C
Figure 8
The effec
etter by plot
nd X
F
= 7%,
dding 10% C
C
2
H
4
increase
8.8 Experi
C
2
H
4
(
C
2
H
4
(
ct of C
2
H
4
a
tting T
ign
ver
, as shown
C
2
H
4
in n-C
es to 20% in
imental T
ign
’
(○), 80% n-C
▲) flames.
addition on
rsus the volu
in Figure 8
C
12
H
26
, while
n the fuel ble
s of C
2
H
4
( ♦
C
12
H
26
/ 20%
the ignition
ume fraction
.9. It can b
e T
ign
is redu
nd.
♦), n-C
12
H
26
(
% C
2
H
4
(▪), a
n of n-C
12
H
n of C
2
H
4
in
be seen that
uced by 20 K
( ▲), 90% n-
and 50% n-
H
26
flames c
the fuel ble
t T
ign
does n
K when the
-C
12
H
26
/ 10%
-C
12
H
26
/ 50%
can be visua
ends at X
F
=
not change w
concentrati
146
%
%
alized
3.5%
when
on of
sh
H
in
d
Figure 8
The sensi
hown in Figu
H
2
/CO and C
n the fuel b
ominant reac
8.9 Experi
80% n
versus
itivity of T
ig
ure 8.10 for
C
1
-C
3
hydroc
blend, R1 an
ctions for C
2
imental T
ign
’
n-C
12
H
26
/ 20
mole fractio
gn
on kinetic
X
F
= 6% an
carbon kinet
nd R2 beco
2
H
4
oxidatio
’s of C
2
H
4
, n
0% C
2
H
4
, an
ons of C
2
H
4
a
cs for n-C
12
H
nd K = 130 s.
ics. Note th
ome more im
on as shown
n-C
12
H
26
, 90
nd 50% n-C
at X
F
= 3.5 %
H
26
and n-C
.
-1
T
ign
was
hat by increa
mportant be
earlier in Fig
0% n-C
12
H
26
C
12
H
26
/ 50%
% and X
F
= 7
C
12
H
26
/C
2
H
4
determined
asing the C
2
ecause they
gure 8.4
/ 10% C
2
H
C
2
H
4
flame
%.
blends flam
to be sensiti
2
H
4
concentr
are the pri
147
H
4
,
es
mes is
ive to
ration
imary
an
N
C
th
ef
Figure 8
Figure 8.1
nd K = 130 s
N
2
with the
C
2
H
4
/n-C
12
H
2
he diffusion
ffects on ign
8.10 Logar
coeffi
flame
11 depicts th
s.
-1
It can be
increasing
26
blends fla
of n-C
12
H
26
nition when a
rithmic sen
cients comp
es with X
F
= 6
he sensitivity
e seen that T
of C
2
H
4
co
ames with 1:
6
, while the d
adding 50%
nsitivity coef
puted for n
6%.
y of T
ign
on b
T
ign
becomes
oncentration
9 and 1:4 vo
diffusivities
C
2
H
4
in the
fficients of
n-C
12
H
26
an
binary diffus
s more sensit
in the fuel
olume ratios
of n-C
12
H
26
e fuel blend.
T
ign
on r
nd n-C
12
H
26
/
sion coeffici
tive to the d
l blend. Fo
s, T
ign
is larg
6
and C
2
H
4
h
C
2
H
4
transp
reaction rat
/C
2
H
4
blend
ients for X
F
diffusion of C
or n-C
12
H
26
gely controlle
have
compar
ports much f
148
.
te
ds
= 6%
C
2
H
4
-
, and
ed by
rative
faster
th
h
8
fl
h
han n-C
12
H
26
igher C
2
H
4
c
Figure 8
C
2
H
4
is a
.12 depicts
lames of n-C
as been con
6
to the igni
concentration
8.11 Logar
coeffi
flame
a main const
the concent
C
12
H
26
and C
nsumed when
ition kernel r
ns.
rithmic sens
cients comp
es with X
F
= 6
tituent of n-
tration profi
C
2
H
4
/n-C
12
H
n the C
2
H
4
c
resulting in
sitivity coeff
puted for n
6%.
-C
12
H
26
deco
les of n-C
12
H
26
blends. A
concentratio
a notable re
fficients of T
n-C
12
H
26
an
omposition
2
H
26
and C
2
As seen in F
on reaches it
eduction of T
T
ign
on bina
nd n-C
12
H
26
/
through s
2
H
4
at the ig
Figure 8.12,
ts peak in th
T
ign
, especia
ary diffusio
/C
2
H
4
blend
scissions. F
gnition state
94% of n-C
he ignition k
149
ally at
on
ds
Figure
es for
C
12
H
26
kernel
fo
w
b
co
re
fu
b
or n-C
12
H
26
with C
2
H
4
ad
efore reachin
onsumption
esulting thus
uel and reac
efore reachin
Figure 8
flames whil
ddition abov
ng the igniti
of C
2
H
4
is
s in a minor
cts at an ear
ng the igniti
8.12 C
2
H
4
and n
(―) n
20% C
le there are n
e 10%. The
ion kernel fo
balanced b
effect on T
ig
rly stage, re
on kernel, fa
and n-C
12
H
n-C
12
H
26
/C
2
H
n-C
12
H
26
, (∙∙∙)
C
2
H
4
, (‒ ‒ ‒)
no peaks of
e concentrat
or 10% C
2
H
4
y its produc
gn
. For C
2
H
4
sulting in n
acilitating th
H
26
concentra
H
4
blends fla
) 90% n-C
12
H
50% n-C
12
H
C
2
H
4
in the
tion of C
2
H
4
4
in the fuel b
ction from t
4
addition ab
notable amou
hus ignition.
ation profile
ames at igni
H
26
/ 10% C
2
H
26
/ 50% C
2
H
e oxidation o
4
does not c
blend, demo
the n-C
12
H
2
bove 10%, C
unts of C
2
H
es computed
ition state w
2
H
4
, (– ∙ –)80
H
4
.
of the fuel b
hange notic
onstrating tha
6
decompos
C
2
H
4
serves a
H
3
and H rad
for n-C
12
H
2
with X
F
= 6%
0% n-C
12
H
26
150
blends
eably
at the
sition,
as the
dicals
26
%.
6
/
151
8.5 Concluding Remarks
Ignition temperatures of non-premixed flames of C
2
H
4
, n-C
12
H
26
, and n-C
12
H
26
/C
2
H
4
blends were measured in the counterflow configuration under atmospheric pressure, a
constant strain rate of 130 s,
-1
and an unburned fuel/N
2
mixture temperature of 453 K.
Laser Doppler Velocimetry was used to determine the local strain rate on the fuel side.
The USC Mech II and the NUI Mech consisting were used to simulate the ignition of
non-premixed C
2
H
4
flames. JetSurF 1.0 was used in the simulations of flames of n-
C
12
H
26
and n-C
12
H
26
/C
2
H
4
blends. The computed ignition temperatures by NUI Mech
were found to be approximately 50 K lower compared to USC Mech II, and in close
agreement with the data for C
2
H
4
flames. JetSurF 1.0 predicts well the data for n-C
12
H
26
flames while the discrepancy between numerical results and the data becomes larger with
the increasing C
2
H
4
concentration.
10% C
2
H
4
addition in the n-C
12
H
26
was determined not to have a measurable effect on
ignition. However, the ignition temperature is reduced by 20 K when the concentration
of C
2
H
4
increases to 20% in the fuel blend, because C
2
H
4
results in notable amounts of
C
2
H
3
and H radicals before reaching the ignition kernel, facilitating thus ignition.
Sensitivity analyses on kinetics and diffusion revealed that the ignition temperature is
primarily sensitive to the diffusivity of n-C
12
H
26
as well as H
2
/CO and C
1
-C
3
small
hydrocarbon kinetics for n-C
12
H
26
and n-C
12
H
26
/C
2
H
4
flames. With increasing C
2
H
4
concentration in the fuel blend, the diffusion of C
2
H
4
affects ignition more notably.
Furthermore, the ignition temperature was found to be more sensitive to
152
C
2
H
4
+ OH → C
2
H
3
+ H
2
O and C
2
H
3
+ O
2
→ CH
2
CHO + O with increasing C
2
H
4
addition.
8.6 References
[1] F.N. Egolfopoulos, P.E. Dimotakis, Proc. Combust. Inst. 27 (1998) 641-648.
[2] F.N. Egolfopoulos, P.E. Dimotakis, Combust. Sci. And Tech. 156 (2000) 173-199.
[3] F.N. Egolfopoulos, P.E. Dimotakis, Combust. Sci. And Tech. 162 (2001) 19-35.
[4] G. Yu, C.K. Law, C.K. Wu, Combust. Flame 63 (1986) 339–347.
[5] M.J. Brown, I.C. Mclean, D.B. Smith, S.C. Taylor, Proc. Combust. Inst. 26 (1996)
875–881.
[6] Y.F. Dong, C.M. Vagelopoulos, G.R. Spedding, F.N. Egolfopoulos, Proc.
Combust. Inst. 29 (2002) 1419–1426.
[7] T. Hirasawa, C.J. Sung, A. Joshi, Z. Yang, H. Wang, C.K. Law, Proc. Combust.
Inst. 29 (2002) 1427-1434.
[8] F. Halter, C. Chauveau, N. Djebai’li-Chaumeix, I. Gökalp, Proc. Combust. Inst. 30
(2005) 201–208.
[9] J. Natarajan, T. Lieuwen, J. Seitzman, Combust. Flame 151 (2007) 104–119.
[10] C.L. Tang, J.J. He, Z.H. Huang, C. Jin, J.H. Wang, X.B. Wang, H.Y. Miao, Int. J.
Hydrogen Energy 33 (2008) 7274–7285.
[11] W. Liu, A.P. Kelley, C.K. Law, Combust. Flame 157 (2010) 1027-1036.
[12] O. Park, P.S. Veloo, N. Liu, F.N. Egolfopoulos, Proc. Combust. Inst. 33 (2011)
887-894.
153
[13] C.G. Fotache, T.G. Kreutz, C.K. Law, Combust. Flame 110 (1997) 429–440.
[14] J.Y.D. Trujillo, T.G. Kreutz, C.K. Law, Combust. Sci. Technol. 127 (1997) 1–27.
[15] C.G. Fotache, Y. Tan, C.J. Sung, C.K. Law, Combust. Flame 120 (2000) 417–426.
[16] S.K. Aggarwal, O. Awomolo, K. Akber, Int. J. Hydrogen Energy 36 (2011)
15392-15402.
[17] N.M. Marinov, W.J. Pitz, C.K. Westbrook, A.M. Vincitore, M.J. Castaldi, S.M.
Senkan, Combust. Flame 114 (1998) 192–213.
[18] Hai Wang, Xiaoqing You, Ameya V. Joshi, Scott G. Davis, Alexander Laskin,
Fokion Egolfopoulos & Chung K. Law, USC Mech Version II. High-Temperature
Combustion Reaction Model of H2/CO/C1-C4 Compounds.
http://ignis.usc.edu/USC_Mech_II.htm, May 2007.
[19] E.L. Petersen, D.M. Kalitan, S. Simmons, G. Bourque, H.J. Curran, J.M. Simmie,
Proc. Combust. Inst. 31 (2007) 447–454.
[20] G. Bourque, D. Healy, H. J. Curran, C. Zinner, D. Kalitan, J. de Vries, C. Aul, E.
Petersen, Proc. ASME Turbo Expo. 3 (2008) 1051–1066.
[21] E. Ranzi, A. Frassoldati, S. Granata, T. Faravelli, Ind. Eng. Chem. Res. 44 (2005)
5170–5183.
[22] X. You, F.N. Egolfopoulos, H. Wang, Proc. Combust. Inst. 32 (2009) 403–410.
[23] C.K. Westbrook, W.J. Pitz, O. Herbinet, H.J. Curran, E.J. Silke, Combust. Flame
156 (2009) 181–199.
154
[24] "Chemical-Kinetic Mechanisms for Combustion Applications", San Diego
Mechanism web page, Mechanical and Aerospace Engineering (Combustion
Research), University of California at San Diego (http://combustion.ucsd.edu).
[25] B. Sirjean, E. Dames, D. A. Sheen, X.-Q. You, C. Sung, A. T. Holley, F. N.
Egolfopoulos, H. Wang, S. S. Vasu, D. F. Davidson, R. K. Hanson, H. Pitsch, C. T.
Bowman, A. Kelley, C. K. Law, W. Tsang, N. P. Cernansky, D. L. Miller, A.
Violi, R. P. Lindstedt, A high-temperature chemical kinetic model of n-alkane
oxidation, JetSurF version 1.0, September 15, 2009
(http://melchior.usc.edu/JetSurF/Version1_0/Index.html).
[26] B. Varatharajan, F.A. Williams, J. Propul. Power 18 (2) (2002) 344– 351.
155
Chapter 9
Flame Ignition Studies of Conventional and Alternative Jet Fuels
9.1 Introduction
Jet fuel chemical compositions can vary in principle within limits though to meet the
stringent specifications associated with air-breathing propulsion both civilian and
military. In general, the composition is dominated by n-, iso-, and cyclo-alkanes as well
as aromatics (e.g., [1]). Due to the economic and environmental challenges, alternative
jet fuels produced via Fischer-Tropsch (FT) process achieved considerable interest in
recent years, which typically consist solely of n- and iso-alkanes without aromatics, such
for example S-8 made from natural gas and a Shell-GTL made from a gas to liquid (GTL)
process (e.g., [2]).
The ignition behavior of those fuels is essential to be reproduced when developing
surrogates [3]. However, fundamental studies on ignition characteristics of practical jet
fuels are relatively limited.
156
Few studies have considered flame properties of JP-8 (e.g., [2-13]), while flame
studies on JP-7, S-8, and Shell-GTL are notably scant (e.g., [2,3,5,6,13-17]). Ji et al. [2]
reported laminar flame speeds and extinction limits of premixed and non-premixed
flames of JP-7, JP-8, S-8, and Shell-GTL, and Holley et al. [3] determined experimentally
extinction and ignition limits of JP-7 and JP-8 in the counterflow configuration. Humer
et al. [4] studied experimentally and numerically the ignition and extinction of laminar
non-premixed JP-8 flames and three possible surrogates were examined. Gokulakrishnan
et al. [5] measured ignition delay times of JP-7, JP-8, and S-8 in a flow reactor between
900 K and 1200 K. Kahandawala et al. [6] reported the similar ignition delays of JP-8
and S-8 in shock tube measurements. Vasu et al. [7] reported the first gas-phase shock
tube ignition delay time data for JP-8 and compared with computed results using several
kinetic models [8-12]. Lenhert et al. [13] studied the pre-ignition behavior of JP-7 and
JP-8 in a pressurized flow reactor for a temperature regime of 600-800 K. Convery et al.
[14] studied the extinction limits of JP-7 measured in an opposed jet burner. Puri et al.
[15] computationally studied the ignition delay times of cracked JP-7 in homogeneous
mixtures. Kumar et al. [16] reported laminar flame speeds and extinction limits of
premixed flames of S-8. Dooley et al. [17] studied the combustion properties of S-8 in a
variable pressure flow reactor, a shock tube, and a counterflow burner.
Based on the considerations, the present work aims to provide archival experimental
data on ignition of non-premixed flames of JP-7, JP-8, S-8, and Shell-GTL, assessing the
variations of ignition propensity on fuel compositions and achieve a better understanding
of their high temperature chemistry property. Similar measurements for n-C
10
H
22
were
157
made and together with the ignition data of n-C
12
H
26
from Chapter 8 were compared
against those obtained for the jet fuels, owing to their similar molecular weight with the
jet fuels and being thus candidate surrogate components (e.g., [10]).
9.2 Experimental Approach
Experiments were performed in the counterflow configuration with nozzle diameter
of 22 mm and a nozzle separation distance of 22 mm. Under all experimental conditions,
T
u
= 453 K and K = 130 s.
-1
All liquid fuels were fully pre-vaporized and mixed with N
2
before the gaseous mixture was injected into the counterflow burner. T
ign
’s were
measured for a wide range of fuel to N
2
mass ratio, (Fuel/N
2
)
mass
, between 0 and 50%.
Two conventional jet fuels and two alternative jet fuels were tested. JP-7 (POSF -
3327), JP-8 (POSF - 3773), S-8 (POSF-4734), and Shell-GTL (POSF - 5172) were
provided by the Air Force Research Laboratory (AFRL) with detailed compositions [2]
illustrated in Figure 9.1 and documented in Table 9.1; “POSF” is the identification
number provided by AFRK. JP-7 and JP-8 contain mainly 67.9% and 57.2% n- and iso-
alkanes on a per mass basis respectively, while S-8 and Shell-GTL consist of almost
entirely n- and iso-alkanes. The aromatic components in JP-8 are approximately 13.5%
of the total mass of the fuel, while those concentrations in JP-7 are much lower with a
mass fraction of 0.7%. S-8 and Shell-GTL contain no aromatics. Cyclo-alkanes are the
secondary major components in JP-7 and JP-8, and those contained in JP-7 is slightly
higher than those in JP-8. Note that Shell-GTL contains 0.8% cyclo-alkanes and S-8 are
less than 0.2%.
Figure 9
9.1 Comp
and S
parison of f
Shell-GTL on
fuel composit
n a per mass
tion distribu
basis.
ution for JP P-7, JP-8, S-8
158
8,
159
Table 9.1 Jet fuel properties and detailed composition on a per mass basis.
Conventional Jet Fuels Alternative Jet Fuels
JP-7 JP-8 S-8 Shell-GTL
Identification number POSF-3327 POSF-3773 POSF-4734 POSF-5172
Approximate formula C
12
H
25
C
11
H
21
C
10
H
22.7
Fuel compositions
Alkanes (n-+iso-) 67.9 57.2 99.7 99.0
Cyclo-alkanes 21.2 17.4 <0.2 0.8
Dicyclo-alkanes 9.4 6.1 <0.1 <0.1
Tricyclo-alkanes 0.6 0.6 <0.1 <0.1
Alkyl-benzenes 0.7 13.5 <0.1 <0.1
Others 0.2 5.2 <0.1 <0.1
n-C
10
H
22
was tested under the same conditions while ignition data of n-C
12
H
26
flames
were reported in Chapter 8 and were also used herein for comparison.
The approximate molecular formula for JP-7, JP-8, and S-8 are listed in Table 9.1.
The molecular weight of Shell-GTL is not available and it is assumed to be close to that
of S-8 given that their chemical compositions are similar (Tim Edwards, private
communications, 2008), but with different n- and iso-alkane distributions as it will be
discussed below.
The experimental uncertainties were estimated to be 25 K using the methodology
described in Chapter 2 caused primarily by the temperature correction due to heat transfer.
160
As all fuels were tested against the same hot boundary conditions, their relative
uncertainties were within 3 K thus their relative performance can be assessed.
9.3 Modeling Approach
JetSurF 2.0 [18] was adopted in the present study for modeling the high temperature
oxidation chemistry of n-dodecane, methylcyclohexane, and toluene, and it has been
validated in previous studies [19,20].
9.4 Results and Discussion
Experimentally determined T
ign
’s of non-premixed flames of JP-7, JP-8, S-8, and
Shell-GTL are shown in Figure 9.2 for a wide range of Fuel to N
2
mass ratio varied
between 0 and 50%. Figure 9.2 depicts also the experimental data of n-C
12
H
26
reported
in Chapter 8 and the experimental T
ign
’s of n-C
10
H
22
measured at the same conditions.
JP-7 and JP-8 exhibit the highest T
ign
’s among all fuels, including alternative fuels, n-
C
10
H
22
, and n-C
12
H
26
. n-C
10
H
22
flames are the easiest to be ignite compared to all jet
fuels. Flames of S-8 and Shell-GTL are harder to ignite than those of n-C
10
H
22
, while
T
ign
’s of S-8 flames are higher than those of Shell-GTL flames and are similar to those of
n-C
12
H
26
flames.
(F
(F
T
es
la
pr
in
Figure 9
It is of in
Fuel/N
2
)
mass
Fuel/N
2
)
mass
This is reason
specially for
argely the i
ropensity of
The oxida
n the previo
9.2 Expe
ratio
Shell
nterest to no
< 30%, w
> 30% and s
nable becau
r heavy fuel
gnition proc
f JP-8 that is
ation kinetic
ous chapter
erimentally d
at T
u
= 453 K
-GTL ( ▲), n
ote that T
ign
’
while JP-8
such differen
use for lower
ls, while wit
cess as it i
caused by th
cs of differen
s, which re
determined
K and K = 1
n-C
10
H
22
( ●)
’s of JP-7 a
8 exhibits
nces increase
r (Fuel/N
2
)
m
th increasing
s from the
he presence
nt chemical
eveals that
T
ign
’s as a f
30 s
-1
of JP-7
), and n-C
12
H
are slightly h
s lower
e with the in
mass
’s, ignitio
g (Fuel/N
2
)
m
reduced, co
of aromatics
class neat c
iso-alkanes
function of
7 ( △), JP-8
H
26
/air ( ◆) fl
higher than
ignition
ncreasing fue
on is diffusi
mass
chemica
ompared to
s [3,4].
components
are much
Fuel/N
2
mas
( ◇), S-8 ( ■
ames.
those of JP
propensity
el concentrat
onally contr
l kinetics co
the other
was investi
harder to i
161
ss
),
P-8 at
for
tions.
rolled
ontrol
fuels,
gated
ignite
162
comparing to n-alkanes [21] and cyclo-alkanes have similar or slightly higher ignition
propensity compared to n-alkanes with comparable carbon numbers [19]. During the
oxidation of iso-alkanes, the concentrations of methyl, propene, and allyl in the ignition
kernel are higher compared to n-alkanes, while the concentration of vinyl radical is
lower. As known in Chapter 6, methyl, propene and allyl are relatively stable and
contribute to chain termination reactions thus inhibiting ignition. Cyclohexane and
methyl-, ethyl-, n-propyl-, and n-butylcyclohexane, have lower or similar ignition
propensity as n-hexane due to the cyclohexyl radical supplied by the fuel subsequently
produces large amount of H radical and other reactive radicals, such as vinyl,
significantly facilitating ignition. Ignition studies of aromatics flames have not been
included in the present study, while their ignition behaviors were predicted by using a
reliable kinetic model, JetSurF 2.0, and shown in Figure 9.3. 13.5 % toluene per mass
basis in n-C
12
H
26
/toluene mixture effectively reduced T
ign
of n-C
12
H
26
by approximately
13 K. Such effect of aromatic components on ignition is more significant than that of
cyclo-alkanes. A flame of 17.4% methylcyclohexane and 82.6% n-C
12
H
26
mixture ignites
at 2 K lower than n-C
12
H
26
flames and shown also in Figure 9.3.
pr
co
G
m
in
in
co
Figure 9
As expec
resence of i
ontains 45%
GTL and S-8
molecular we
n higher fuel
n Shell-GTL
ontain 45%
9.3 Comp
temp
at T
u
C
12
H
n-C
12
ted, JP-7 an
iso-alkanes a
% of iso-alka
flames are
eights of both
l diffusivity
L and S-8 ten
[26] and 80
puted H m
eratures usin
u
= 453 K an
26
+ 17.4 % m
2
H
26
+ 13.5 %
nd JP-8 flam
and/or arom
anes [6] on a
lower and si
h Shell-GTL
that tends to
nds to inhibit
0% iso-alkan
mass fractio
ng JetSurF 2
d K = 130 s.
methylcycloh
% toluene per
mes are harde
matics in thei
a per mass b
imilar respec
L and S-8 are
o facilitate ig
t ignition co
nes [6,17] o
on response
2.0 for a fue
.
-1
( ― ) n-
hexane per m
r mass basis.
er to ignite
ir compositio
basis. On th
ctively to th
e smaller co
gnition, whil
ompared to n
on a per ma
e to the h
l mole fracti
-C
12
H
26
, ( ---
mass basis, (
.
than n-C
12
H
on that inhib
he other hand
hose of n-C
12
ompared to n
le the presen
n-C
12
H
26
. Sh
ass basis resp
ot boundar
ion, X
F
= 6 %
- ) 82.6 % n
∙ ‒ ∙ )86.5 %
H
26
flames d
bit ignition;
d, T
ign
’s of S
2
H
26
flames.
n-C
12
H
26
resu
nce of iso-alk
hell-GTL an
pectively. T
163
ry
%
n-
%
due to
JP-8
Shell-
The
ulting
kanes
d S-8
Thus,
164
comparing S-8 and n-C
12
H
26
flames the high iso-alkane content of S-8 counterbalances
the effect of diffusivity with respect to ignition propensity. Comparing the Shell-GTL
and S-8 flames, the higher iso-alkane content of S-8 results in reduced ignition propensity
compared to Shell-GTL.
9.5 Concluding Remarks
Ignition temperatures of non-premixed JP-7, JP-8, S-8, and Shell-GTL flames were
determined experimentally as functions of fuel/air mass ratio at an unburned gaseous
mixture temperature of 453 K and a local strain rate of 130 s
-1
measured on the fuel side.
The simulation results using JetSurF 2.0 identified that the presence of aromatics notably
reduces the ignition propensity and cyclo-alkanes are slightly promoting ignition. The
experimental results revealed that the ignition temperatures of JP-7 and JP-8 flames are
higher than n-dodecane, S-8, Shell-GTL, and n-decane flames, while n-decane exhibits
the highest ignition propensity compared with all fuels. Shell-GTL was found to ignite
harder than n-decane and easier than S-8, while S-8 has the similar ignition temperatures
as n-dodecane. This effect is caused by both the fuel diffusivity and differences in the
oxidation kinetics of those fuel components of different chemical classes.
9.6 References
[1] T. Edwards, J. Propul. Power 19 (6) (2003) 1089–1107.
[2] C. Ji, Y.L. Wang, F.N. Egolfopoulos, J. Propul. Power 27 (2011) 856-863.
165
[3] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, T. Edwards, Proc.
Combust. Inst. 31 (2007) 1205–1213.
[4] S. Humer, A. Frassoldati, S. Granata, T. Faravelli, E. Ranzi, R. Seiser, K. Seshadri,
Proc. Combust. Inst. 31 (2007) 393-400.
[5] P. Gokulakrishnan, G. Gaines, M.S. Klassen, R.J. Roby, 43
rd
AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibits 8 -11 July
2007, Cincinnati, OH.
[6] M.S.P. Kahandawala, M.J. DeWitt, E. Corporan, S.S. Sidhu, Energy Fuels 22
(2008) 3673-3679.
[7] S.S. Vasu, D.F. Davidson, R.K. Hanson, Combust. Flame 152 (2008) 125–143.
[8] M.A. Mawid, T.W. Park, B. Sekar, C. Arana, 38
th
Joint Propulsion Conference &
Exhibits, Indianapolis, IN, AIAA, July 2002, pp. 2002-3876.
[9] R.P. Lindstedt, L.Q. Maurice, J. Propul. Power 16 (2) (2000) 187-195.
[10] P. Dagaut, M. Cathonnet, Prog. Energy Combust. Sci. 32 (2006) 48-92.
[11] E. Ranzi, A. Frassoldati, S. Granata, T. Faravelli, Ind. Eng. Chem. Res. 44 (14)
(2005) 5170-5183.
[12] H.R. Zhang, E.G. Eddings, A.F. Sarofim, Proc. Combust. Inst. 31 (2007) 401–409.
[13] D.B. Lenhert, D.L. Miller, N.P. Cernansky, Combust. Sci. Technol. 179 (5) (2007)
845-861.
[14] J.L. Convery, G.L. Pellett, W.F. O’Brien, L.G. Wilson, 41
st
AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibits 10 -13 July
2005, Tucson, Arizona.
166
[15] P. Puri, R. Ma, J. Choi, V. Yang, Combust. Flame 142 (4) (2005) 454-457.
[16] K. Kumar, C.J. Sung, X. Hui, 47th AIAA Aerospace Sciences Meeting, AIAA
2009-991, Orlando, Florida, 5–8 January, 2009.
[17] S. Dooley, S.H. Won, S. Jahangirian, Y. Ju, F.L. Dryer, H. Wang, M.A.
Oehlschlaeger, Combust. Flame 159 (2012) 3014–3020.
[18] H. Wang, E. Dames, B. Sirjean, D. A. Sheen, R. Tangko, A. Violi, J. Y. W. Lai, F.
N. Egolfopoulos, D. F. Davidson, R. K. Hanson, C. T. Bowman, C. K. Law, W.
Tsang, N. P. Cernansky, D. L. Miller, R. P. Lindstedt, A high-temperature
chemical kinetic model of n-alkane (up to n-dodecane), cyclohexane, and methyl-,
ethyl-, n-propyl and n-butyl-cyclohexane oxidation at high temperatures, JetSurF
version 2.0, September 19, 2010 (http://melchior.usc.edu/JetSurF/JetSurF2.0).
[19] N. Liu, C. Ji, F.N. Egolfopoulos, Proc. Combust. Inst. 34 (2012) doi:
10.1016/j.proci.2012.05.104
[20] C. Ji, E. Dames, H. Wang, F.N. Egolfopoulos, Combust. Flame 159 (3) (2012)
1070-1081.
[21] N. Liu, S.M. Sarathy, C.K. Westbrook, F.N. Egolfopoulos, Proc. Combust. Inst.
(2012), http://dx.doi.org/10.1016/j.proci.2012.05.040.
[22] E. Corporan, T. Edwards, L. Shafer, M.J. DeWitt, C. Klingshirn, S. Zabarnick, Z.
West, R. Striebich, J. Graham, J. Klein, Energy Fuels 25 (2011) 955-966.
167
Chapter 10
Mixing Rules of Ignition Phenomena
10.1 Introduction
Practical fuels are composed of a number of neat components and involve complex
oxidation kinetics. Experimental studies on the combustion behavior of all possible fuel
blends are not feasible. Therefore, developing efficient and sufficiently accurate rules
that could allow for the prediction of the combustion properties of fuel blends based on
their composition is highly desirable.
Ji and Egolfopoulos [1] developed a mixting rule to estimate the laminar flame speeds
of binary liquid fuel mixtures by knowing the adiabatic flame temperatures and laminar
flame speeds of individual components similar to the mixing rule reported by Hirasawa et
al. [2]. Core and coworkers [3-5] reported a mixing rule for the estimation of species
concentrations in flames of fuel mixtures. Cheng and Oppenheim [6] proposed a
heuristic formula to estimate ignition delay times of methane/hydrogen/oxygen mixtures
behind a reflected shock tube.
168
The present work aims to study the correlation of ignition behavior of fuels blends
and their neat components in homogeneous systems, and extend to the development of
mixing rules to estimate the ignition states of counterflow flames.
10.2 A Mixing Rule in Homogeneous Systems
Ignition in homogeneous systems is studied typically in shock tubes or rapid
compression machines, and ignition delay times are determined experimentally and
numerically. In order to estimate the ignition delay times, , of binary fuel blends, an
analysis similar to that reported by Cheng and Oppenhem [6] was adopted herein.
The Arrhenius expression for ignition delay is
exp
/ ,
where E
a
is the overall activation energy, R the universal gas constant, T the prevailing
temperature, and A, a, and b are constants. The quantities in the brackets denote
concentration in mole/cm
3
. For a given fuel concentration and by keeping the oxygen
concentration constant, the expression can be simplified to ~exp /RT , indicating
that τ is controlled by E
a
and T.
For a binary fuel blend, the compositions are specified as:
∑
where X
i
is the mole fraction of the ith component in the fuel/oxidizer mixture.
Assuming ,
∑ ,
and ∏
, then the ignition delay times of the
mixture can be expressed as
exp
,
/ .
co
co
ra
4
re
By furthe
orrelation of
The com
orresponding
atio of 50%
0% : 20% p
espectively.
er assuming
f ignition del
mputations o
g estimation
: 50% per v
per volume
g ∑
lay time of a
of the ign
ns for those
volume and
at 1 atm a
, , a binary fuel
nition delay
fuel blends
the fuel ble
are shown in
(a)
, and
l mixture, τ
m
ys using U
s of H
2
/CH
4
end of H
2
/CO
n Figures 1
d ,
m
, can be then
USC_Mech_
4
, CH
4
/C
3
H
8
O/CH
4
with
0.1(a), 10.1
[6], then
n derived as
_II [7] and
with a blen
a ratio of 4
(b), and 10
169
n the
d the
nding
40% :
0.1(c),
Figure 1 10.1 Comp
delay
(b) 50
= 1 :
putations us
y times of bin
0% CH
4
/ 50
1, per volum
(b)
(c)
sing USC_M
nary fuel ble
0% C
3
H
8
, (c)
me.
Mech_II and
ends at = 1
) 20% CH
4
/
d estimation
1. (a) 50% C
80% (H
2
+ C
ns of ignitio
CH
4
/ 50% H
CO), H
2
: CO
170
on
H
2
,
O
1
ch
th
st
th
0.3 A Mix
In counter
haracteristic
he ignition s
/
treamline an
hen as a func
Figure 1
xing Rule in
rflow flames
time of the
strain rate, K
at the igniti
nd x is the di
ction of hot b
10.2 Igniti
CH
4
;
best f
n Counterflo
s, the recipro
e transport o
K
ign
, can be
ion state, in
istance from
boundary tem
ion response
( ▲) 75% C
fits of compu
ow Flames
ocal value o
of the reactan
e defined as
which u is t
m the burner
mperatures,
e of non-pre
CH
4
; ( ■) 50%
utational valu
of the local s
nts into the
s the maxim
the axial vel
exit on the f
T.
emixed meth
% CH
4
; ( ◇) 1
ues.
train rate, 1/
ignition ker
mum local s
locity along
fuel side. K
hane flames
10% CH
4
in
/K, represent
rnel. As a r
strain rate the center o
K
ign
can be pl
s. ( ◆) 100%
N
2
. Lines ar
171
ts the
result,
of the
lotted
%
re
Figure 1
Figure 1
10.3 Igniti
( ■) 5
comp
10.4 Igniti
flame
blend
ion response
50% C
3
H
8
;
putational va
ion respons
es. ( ▲) 75%
d in N
2
. Lines
e of non-prem
( ◇) 10%
alues.
se of non-
% fuel blend;
s are best fits
mixed propan
C
3
H
8
in N
2
-premixed m
; ( ■) 50% fu
s of computa
ne flames. ( ▲
2
. Lines are
methane/pro
fuel blend; (
ational values
▲) 75% C
3
H
8
e best fits o
opane blend
◇) 10% fue
s.
172
8
;
of
ds
el
173
The ignition response of non-premixed flames of methane, propane, and
mechane/propane blends with volume ratio of 1:1 were computed and shown in Figures
10.2, 10.3, and 10.4, respectively. Simple phenomenological analysis reveales that
1/
/ , where E
a
is the overall activation energy, R the ideal gas
constant, and T the hot boundary temperature. Herein ,,
, is a function of
fuel mole fraction, X
F
, pressure, P, and activation energy, E
a
. For a given pressure and
fuel mole fraction, 1/
~
/ . It can be seen that K
ign
is controlled by
E
a
and T. E
a
can be obtained using linear regression method by the best fits of
computional results. Figure 10.2 shows that E
a
’s of CH
4
flames at the ignition state are
39.94 Kcal/mol, 43.87 Kcal/mol, 44.27 Kcal/mol, and 43.14 Kcal/mol for fuel mole
fractions of 10 %, 50 %, 75 %, and 100 %, respectively. E
a
’s of C
3
H
8
flames and CH
4
/
C
3
H
8
blend flames are shown in Figure 10.3 and 10.4.
In order to estimate the K
ign
of non-premixed flames of binary fuel blends, the
methodology developed by Ji et al. [1] was adopted. The K
ign
of the ith component in the
fuel blend can be expressed as ln 1/
,
~ ln
,
/ . The assumptions of A
m
and E
a,m
similar to that in homogeneous systems were adopted, which is ∏
and ,
∑ ,
. Combining the above equations, the ignition strain rate of the
fuel blend, K
ign,m
, can be derived as:
ln 1/
~ ln ,
/
Substituting the K
ign
of the ith component into the above equation, then it is expressed:
T
F
o
5
C
co
1
Thus, the corr
The abov
igure 10.5 i
f CH
4
/C
3
H
8
0%, and 10
C
3
H
8
are diff
omputed val
0.7 and were
relation form
ve equation
llustrates tha
blends with
%, respectiv
ferent, the es
lues. The es
e closely to t
ln 1/
mula of K
ign
depicts that
at the compu
h volume ra
vely. Altho
stimations u
stimations o
the compute
of the binary
,
t ,
can
uted and est
atio of 50%
ough the con
sing above e
f A
m
and E
a
ed values, de
(a)
ln1/
y fuel blend
,
be estimate
timated
: 50% at th
ntrolling ign
equation are
,m
are also id
emonstrating
,
becomes:
ed by know
,
’s of non-
he fuel mole
nition chemi
e in close ag
dentified in
g the accepta
wing ,
an
-premixed fl
e faction of
istry of CH
4
greement wit
Figures 10.6
able assumpt
174
nd .
lames
75%,
4
and
th the
6 and
tions.
Figure 1 10.5 Comp
meth
blend
resul
estim
putations an
hane/propane
d; (b) 50% fu
ts; ( ■) estim
mated values.
(b)
(c)
nd estimation
e blends with
uel blend; (c)
mations. Li
ns of K
ign
of
h volume ra
) 10% fuel b
ines are be
f non-premix
atio of 1:1.
lend in N
2
; (
st fits of co
xed flames o
(a) 75% fue
▲) compute
omputed an
175
of
el
ed
nd
Figure 1
Figure 1
10.6 Comp
prem
1:1 a
estim
10.7 Comp
non-p
ratio
resul
putations an
mixed flames
s a function
mations.
putations an
premixed fla
of 1:1 as a
ts; ( ■) estima
nd estimation
of methane/
of fuel mole
nd estimation
ames of me
function of
ations.
ns of overall
/propane ble
e fractions. (
ns of overall
ethane/propa
f fuel mole f
activation en
ends with vo
( ▲) compute
pre-exponen
ane blends
fractions. ( ▲
nergy of non
olume ratio o
ed results; ( ■
ntial factor o
with volum
▲) compute
176
n-
of
■)
of
me
ed
177
10.4 Concluding Remarks
A mixing rule in homogeneous systems was developed and resulted in satisfactory
estimates of ignition delay times of binary fuel blends based on the fuel blend
composition. A mixting rule was subsequently developed for the ignition of counterflow
flames of binary fuel blends flames based also on the fuel blend composition and the
predictions were also close to the actual ignition states computed directly for the fuel
blend flames.
10.5 References
[1] C. Ji, F.N. Egolfopoulos, Proc. Combust. Inst. 33 (2011) 955–961.
[2] T. Hirasawa, C.J. Sung, A. Joshi, Z. Yang, H. Wang, C.K. Law, Proc. Combust.
Inst. 29 (2002) 1427–1434.
[3] J.P. Core, S.M. Skinner, Combust. Flame 87 (1991) 357-364.
[4] J.P. Core, S.M. Skinner, D.W. Stroup, D. Madrzykowski, D.D. Evans, ASME
HTD, New York, 1989, vol. 122, pp. 77-86.
[5] J.P. Core, S.M. Skinner, ASME HTD, New York, 1990, vol. 141, pp. 39-47.
[6] R.K. Cheng, A.K. Oppenheim, Combust. Flame 58 (1984) 125-139.
[7] Hai Wang, Xiaoqing You, Ameya V. Joshi, Scott G. Davis, Alexander Laskin,
Fokion Egolfopoulos & Chung K. Law, USC Mech Version II. High-Temperature
Combustion Reaction Model of H2/CO/C1-C4 Compounds.
http://ignis.usc.edu/USC_Mech_II.htm, May 2007.
178
Chapter 11
Conclusions and Recommendations
11.1 Conclusions
In the present dissertation a study was conducted on the ignition characteristics of
conventional and alternative jet fuels as well as neat and fuel blends of relevance. The
study included experiments and detailed numerical simulations aiming to provide insight
into the physical and chemical processes that control flame ignition for those heavy
hydrocarbons as well as lighter hydrocarbons whose kinetics constitute an important sub-
set of the heavier ones. Ignition limits were investigated in the counterflow configuration
for premixed flames of gaseous fuel blends, and non-premixed flames of conventional
and alternative jet fuels and heavier neat hydrocarbons. All experiments were done in a
properly designed ignition burner system with well-distributed temperature and velocity
radial profiles assuring a quasi one-dimensional configuration that is conforming to the
assumptions of existing opposed-jet numerical codes. The experiments were carried out
at atmospheric pressure and elevated fuel-carrying stream temperatures. The
179
experimental data were validated using a variety of chemical kinetic mechanisms. A
newly developed model for 2,7-dimethyloctane, 2,5-dimethylhexane, 3-methylheptane, n-
octane, and n-decane was introduced herein and an iso-octane model was updated for
better prediction of flame ignition and flame propagation. Rigorous sensitivity analyses
on both chemical kinetics and molecular transport were performed to illustrate the
chemical and physical mechanisms that dominate ignition. Additionally, reaction path
analysis was implemented to provide insight into the consumption pathways of various
fuels at the ignition state.
For both premixed and non-premixed flames, sensitivity analyses revealed that
ignition is notably sensitive to fuel diffusion. Regarding the kinetics, the ignition of
normal, branched, and cyclic alkanes flames were shown to be mostly sensitive to H
2
/CO
and C
1
-C
4
small hydrocarbon chemistry, while some influence was identified from the
fuel related reactions for branched alkanes.
The ignition of premixed gaseous fuel blends flames was carried out also. Due to the
high concentration of H
2
in all fuel blends, ignition temperature was found to be
insensitive to equivalence ratio. C
3
H
8
in the fuel mixture blending with H
2
and CO was
found to produce more methyl radicals compared to CH
4
blending in the same fuel
mixture, exhibiting thus has a lower ignition propensity.
The trend of ignition temperatures of normal alkanes was found to be opposite at low
and high fuel concentrations as a result of the competition between fuel diffusivity and
thermal decomposition. The overall reactivity of branched alkanes was found to decrease
with increasing fuel branching, as expected. Flames of 2,5-dimethylhexane and 2,7-
180
dimethyloctane exhibit similar ignition temperatures even as the influences of fuel
diffusivity and reactivity cancel out. Cyclohexane flames exhibit the highest ignition
propensity among all mono-alkylated cyclohexanes due to its higher diffusivity and
reactivity related to its specific six saturated carbon bond ring structure, releasing more
H, HCO, and C
2
H
3
radicals; moreover, the ignition of mono-alkylated cyclohexane
flames was found to performs as normal hexane. On the other hand, the size of alkyl
group chain (one to four carbons) was found experimentally to have no measurable effect
on ignition, and sensitivity and reaction path analyses showed that this behavior stems
from the competition between fuel diffusivity and reactivity resulted in the distribution of
key cracked species of fuels. The JetSurF 2.0 kinetic model reproduces in general this
trend, except that it predicts a reduced ignition propensity of n-butyl-cyclohexane flames
due to lower concentrations of reactive small species, such as C
2
H
3
.
A binary fuel blends study was carried out for ethylene and n-dodecane blends ratios
of 1 : 1, 1 : 9, and 1 : 4. The promoting effect of ethylene on ignition of n-dodecane
flames was examined and shown that 10% C
2
H
4
addition in the n-C
12
H
26
has no
measurable effect on ignition and the ignition temperature is reduced by 20 K when the
concentration of C
2
H
4
increases to 20% in the fuel blend. Notably different predictions
of ignition temperatures of ethylene flames were observed using USC Mech II, NUI
Mech, and UCSD Mech, respectively, identifying thus potential deficiencies in those
kinetic models. It was found that the large difference in CH
2
CHO chemistry and
uncertainties in the rate coefficients of several key reactions involving CH
2
O result in the
uncertainties in predicting ignition of ethylene flames.
181
The ignition of conventional and alternative jet fuels flames was investigated and the
attendant behaviors were compared to each other. The presence of iso-alkanes reduces
the ignition propensity of JP-7 and JP-8 flames compared to n-dodecane flames. The
absence of aromatic components increased relative reactivity of S-8, and therefore S-8
shown higher ignition propensity than JP-8. Shell-GTL is found to ignite harder than n-
decane and easier than S-8, while S-8 has the similar ignition temperatures as n-
dodecane. Interestingly to note that, T
ign
’s of JP-7 are slightly higher than those of JP-8
at (Fuel/N
2
)
mass
< 30 %, while JP-8 exhibits lower ignition propensity at
(Fuel/N
2
)
mass
> 30% and such differences increase with the increasing fuel concentrations.
This effect is caused by both the fuel diffusivity and differences in the oxidation kinetics
of those fuel components of different chemical classes.
Finally, a quasi-empirical correlation formula of ignition mixing rules was derived for
both homogeneous systems and flames based on the assumption of no kinetic couplings
in the overall oxidation process. A simple phenomenological analysis suggests that for a
given pressure and initial fuel concentration, the ignition strain rate correlates closely
with the overall activation energy and the hot boundary temperature. The ignition strain
rates of binary fuel blends flames could be estimated with satisfactory accuracy by
knowing the ignition strain rates of the neat fuel components and such estimations were
found to be in good agreement with the simulated results.
182
11.2 Recommendations for Future Work
The present work focused largely on the flame ignition of conventional and
alternative jet fuels as well as neat components of relevance. Additionally, studies on
binary fuel blends and multi-components fuel blends were carried out. Such approach is
essential to achieve an understanding of the ignition characteristics of practical fuels.
Some recommendations for additional future work are given below:
Ignition of fuel blends is of fundamental interest in developing and accessing the
validity of detailed kinetic models. However, the investigations on fuels blends are
limited. Considering the large number of hydrocarbons contained in practical fuels, it is
essential to examine the kinetic interactions of main intermediate species of the fuel
blends that may reveal that new important reaction pathways that cannot be identified
from studies of neat fuels.
The present study was varied out under constant local strain rate conditions, focusing
on the fuel concentration effect. Ignition is directly affected by heat loss and the local
strain rate just upstream of the ignition kernel controls the extent of heat and radical
losses from the ignition kernel. Therefore, ignition behavior can differ from low stain
rates to high strain rates that should be considered in the future work.
In practical engines, such as jet turbines, combustion takes place in general under
high pressures. However, a major problem is also the high-altitute relight in which the
prevailing pressure is a fraction of 1 atm. Thus, studies at various pressures are essential.
However, studies for pre-vaportized heavy fuels will be a great challenge given their low
vapor pressure and their relatively low decomposition temperature.
183
Last but not least, ignition of premixed heavy hydrocarbons needs to be investigated
systematically. In this dissertation, on a pilot study has been done for premixed flames of
gaseous hydrocarbons. At present, there are no ignition data for premixed flames of
heavy hydrocarbons. While in engines the fuel and air stream are separated initially, it is
possible that full premixing of fuel and oxygen could take place in various parts of the
engines and such regions could be the most favorable ones to initiate ignition.
184
Bibliography
Annual Energy Outlook (2012): U.S. Energy Information Administration.
Aggarwal S. K., Awomolo O., and Akber K. (2011). Ignition characteristics of
heptane–hydrogen and heptane–methane fuel blends at elevated
pressures. International Journal of Hydrogen Energy, 36(23), 15392-
15402.
Agosta A., Cernansky N. P., Miller D. L., Faravelli T., and Ranzi E. (2004).
Reference components of jet fuels: kinetic modeling and experimental
results. Experimental Thermal and Fluid Science, 28(7), 701-708.
Allen C., Toulson E., Edwards T., and Lee T. (2012). Application of a novel
charge preparation approach to testing the autoignition characteristics of
JP-8 and camelina hydroprocessed renewable jet fuel in a rapid
compression machine. Combustion and Flame, 159(9), 2780-2788.
Andac M. G., and Egolfopoulos F. N. (2007). Diffusion and kinetics effects on the
ignition of premixed and non-premixed flames. Proceedings of the
Combustion Institute, 31(1), 1165-1172.
Battin-Leclerc F., Fournet R., Glaude P. A., Judenherc B., Warth V., Côme G. M.,
and Scacchi G. (2000). Modeling of the gas-phase oxidation of n-decane
from 550 to 1600 K. Proceedings of the Combustion Institute, 28(2), 1597-
1605.
Benson S. W. (1976). Thermochemical Kinetics (second ed.). Wiley, New york.
Bieleveld T., Frassoldati A., Cuoci A., Faravelli T., Ranzi E., Niemann U., and
Seshadri K. (2009). Experimental and kinetic modeling study of
combustion of gasoline, its surrogates and components in laminar non-
premixed flows. Proceedings of the Combustion Institute, 32(1), 493-500.
Bikas G., and Peters N. (2001). Kinetic modelling of n-decane combustion and
autoignition: Modeling combustion of n-decanem. Combustion and Flame,
126(1–2), 1456-1475.
185
Blouch J. D., and Law C. K. (2000). Non-premixed ignition of n-heptane and iso-
octane in a laminar counterflow. Proceedings of the Combustion Institute,
28(2), 1679-1686.
Bourque G., Healy D., Curran H., Zinner C., Kalitan D., Vries J. d., Aul C., and
Petersen E. (2008). Ignition and Flame Speed Kinetics of Two Natural
Gas Blends With High Levels of Heavier Hydrocarbons. ASME
Conference Proceedings, 2008(43130), 1051-1066.
Brown M. J., McLean I. C., Smith D. B., and Taylor S. C. (1996). Markstein
lengths of CO/H2/air flames, using expanding spherical flames.
Symposium (International) on Combustion, 26(1), 875-881.
Buda F., Heyberger B., Fournet R., Glaude P.-A., Warth V., and Battin-Leclerc F.
(2006). Modeling of the Gas-Phase Oxidation of Cyclohexane. Energy &
Fuels, 20(4), 1450-1459.
Burcat A., Farmer R. C., Espinoza R. L., and Matula R. A. (1979). Comparative
ignition delay times for selected ring-structured hydrocarbons. Combustion
and Flame, 36(0), 313-316.
Burcat A., Scheller K., and Lifshitz A. (1971). Shock-tube investigation of
comparative ignition delay times for C1-C5 alkanes. Combustion and
Flame, 16(1), 29-33.
Carlier M., Corre C., Minetti R., Pauwels J. F., Ribaucour M., and Sochet L. R.
(1991). Autoignition of butane: A burner and a rapid compression machine
study. Symposium (International) on Combustion, 23(1), 1753-1758.
Cavallotti C., Rota R., Faravelli T., and Ranzi E. (2007). Ab initio evaluation of
primary cyclo-hexane oxidation reaction rates. Proceedings of the
Combustion Institute, 31(1), 201-209.
Chèze B., Gastineau P., and Chevallier J. (2011). Forecasting world and regional
aviation jet fuel demands to the mid-term (2025). Energy Policy, 39(9),
5147-5158.
Chemical-Kinetic Mechanisms for Combustion Applications. In San Diego
Mechanism web page M. a. A. E. C. R., University of California at San
Diego (http://combustion.ucsd.edu).
Cheng R. K., and Oppenheim A. K. (1984). Autoignition in methanehydrogen
mixtures. Combustion and Flame, 58(2), 125-139.
186
Ciezki H. K., and Adomeit G. (1993). Shock-tube investigation of self-ignition of
n-heptane-air mixtures under engine relevant conditions. Combustion and
Flame, 93(4), 421-433.
Coats C. M., and Williams A. (1979). Investigation of the ignition and combustion
of n-heptane-oxygen mixtures. Symposium (International) on Combustion,
17(1), 611-621.
Colket III M. B., and Spadaccini L. J. (2001). Scramjet fuels autoignition study.
Journal of Propulsion and Power, 17(2), 315-323.
Convery J. L., Pellett G. L., O’Brien W. F., and Wilson L. G. (2005). An
Experimental Study of n-Heptane and JP-7 Extinction Limits in an
Opposed Jet Burner. Paper presented at the 41st Joint Propulsion
Conference and Exhibit AIAA 2005-3776, Tucson, Arizona, 10-13 July.
Cooke J. A., Bellucci M., Smooke M. D., Gomez A., Violi A., Faravelli T., and
Ranzi E. (2005). Computational and experimental study of JP-8, a
surrogate, and its components in counterflow diffusion flames.
Proceedings of the Combustion Institute, 30(1), 439-446.
Gore J. P., Skinner S. M., Stroup D. W., Madrzykowski D., and Evans D. D.
(1989). Structure and radiation properties of large two-phase flames.
Waste Heat Utilization: Presented at the Winter Annual Meeting of the
ASME;San Francisco, CA, USA. 10-15 Dec. vol. 122, pp. 77-86.
Gore J. P., and Skinner S. M. (1990). Radiation properties of laminar and
turbulent methane+acetylene/air diffusion flames. Fundamentals of
Natural Convection - Presented at AIAA/ASME Thermophysics and Heat
Transfer Conference;Seattle, WA, USA. 18-20 Jun. vol. 141, pp. 39-47.
Gore J. P., and Skinner S. M. (1991). Mixing rules for state relationships of
methane and acetylene/air diffusion flames. Combustion and Flame,
87(3–4), 357-364.
Corporan E., Edwards T., Shafer L., DeWitt M. J., Klingshirn C., Zabarnick S.,
West Z., Striebich R., Graham J., and Klein J. (2011). Chemical, Thermal
Stability, Seal Swell, and Emissions Studies of Alternative Jet Fuels.
Energy & Fuels, 25(3), 955-966.
Cox A., Griffiths J. F., Mohamed C., Curran H. J., Pit W. J., and Westbrook C. K.
(1996). Extents of alkane combustion during rapid compression leading to
single-and two-stage ignition. Symposium (International) on Combustion,
26(2), 2685-2692.
187
Crochet M., Minetti R., Ribaucour M., and Vanhove G. (2010). A detailed
experimental study of n-propylcyclohexane autoignition in lean conditions.
Combustion and Flame, 157(11), 2078-2085.
Curran H. J., Gaffuri P., Pitz W. J., and Westbrook C. K. (1998). A
Comprehensive Modeling Study of n-Heptane Oxidation. Combustion and
Flame, 114(1–2), 149-177.
Curran H. J., Gaffuri P., Pitz W. J., and Westbrook C. K. (2002). A
comprehensive modeling study of iso-octane oxidation. Combustion and
Flame, 129(3), 253-280.
Dagaut P. (2002). On the Kinetics of Hydrocarbons Oxidation from Natural Gas
to Kerosene and Diesel Fuel. Physical Chemistry Chemical Physics, 4,
2079-2094.
Dagaut P., and Cathonnet M. (2006). The ignition, oxidation, and combustion of
kerosene: A review of experimental and kinetic modeling. Progress in
Energy and Combustion Science, 32(1), 48-92.
Daley S. M., Berkowitz A. M., and Oehlschlaeger M. A. (2008). A shock tube
study of cyclopentane and cyclohexane ignition at elevated pressures.
International Journal of Chemical Kinetics, 40(10), 624-634.
Darcy D., Mehl M., Simmie J. M., Würmel J., Metcalfe W. K., Westbrook C. K.,
Pitz W. J., and Curran H. J. (2012). An experimental and modeling study
of the shock tube ignition of a mixture of n-heptane and n-propylbenzene
as a surrogate for a large alkyl benzene. Proceedings of the Combustion
Institute, 34, http://dx.doi.org/10.1016/j.proci.2012.06.131.
Davidson D. F., Gauthier B. M., and Hanson R. K. (2005). Shock tube ignition
measurements of iso-octane/air and toluene/air at high pressures.
Proceedings of the Combustion Institute, 30(1), 1175-1182.
Davidson D. F., Haylett D. R., and Hanson R. K. (2008). Development of an
aerosol shock tube for kinetic studies of low-vapor-pressure fuels.
Combustion and Flame, 155(1–2), 108-117.
Davidson D. F., Hong Z., Pilla G. L., Farooq A., Cook R. D., and Hanson R. K.
(2011). Multi-species time-history measurements during n-dodecane
oxidation behind reflected shock waves. Proceedings of the Combustion
Institute, 33(1), 151-157.
188
Davidson D. F., Oehlschlaeger M. A., and Hanson R. K. (2007). Methyl
concentration time-histories during iso-octane and n-heptane oxidation
and pyrolysis. Proceedings of the Combustion Institute, 31(1), 321-328.
Davidson D. F., Oehlschlaeger M. A., Herbon J. T., and Hanson R. K. (2002).
Shock tube measurements of iso-octane ignition times and OH
concentration time histories. Proceedings of the Combustion Institute,
29(1), 1295-1301.
Davis S. G., Joshi A. V., Wang H., and Egolfopoulos F. (2005). An optimized
kinetic model of H2/CO combustion. Proceedings of the Combustion
Institute, 30(1), 1283-1292.
Davis S. G., and Law C. K. (1998). Determination of and Fuel Structure Effects
on Laminar Flame Speeds of C1 to C8 Hydrocarbons. Combustion
Science and Technology, 140(1-6), 427-449.
Dixon-Lewis G. (1968). Flame Structure and Flame Reaction Kinetics. II.
Transport Phenomena in Multicomponent Systems. Proceedings of the
Royal Society A 304 111-135.
Dong Y., Holley A. T., Andac M. G., Egolfopoulos F. N., Davis S. G., Middha P.,
and Wang H. (2005). Extinction of premixed H2/air flames: Chemical
kinetics and molecular diffusion effects. Combustion and Flame, 142(4),
374-387.
Dong Y., Vagelopoulos C. M., Spedding G. R., and Egolfopoulos F. N. (2002).
Measurement of laminar flame speeds through digital particle image
velocimetry: Mixtures of methane and ethane with hydrogen, oxygen,
nitrogen, and helium. Proceedings of the Combustion Institute, 29(2),
1419-1426.
Dooley S., Won S. H., Jahangirian S., Ju Y., Dryer F. L., Wang H., and
Oehlschlaeger M. A. (2012). The combustion kinetics of a synthetic
paraffinic jet aviation fuel and a fundamentally formulated, experimentally
validated surrogate fuel. Combustion and Flame, 159(10), 3014-3020.
DuBois T. G., and Nieh S. (2011). Selection and performance comparison of jet
fuel surrogates for autothermal reforming. Fuel, 90(4), 1439-1448.
Eddings E. G., Yan S., Ciro W., and Sarofim A. F. (2005). FORMULATION OF A
SURROGATE FOR THE SIMULATION OF JET FUEL POOL FIRES.
Combustion Science and Technology, 177(4), 715-739.
189
Edwards T. (2003). Liquid Fuels and Propellants for Aerospace Propulsion:
1903-2003. Journal of Propulsion and Power, 19(6), 1089-1107.
Egoifopoulos F. N., and Dimotakis P. E. (2001). A Comparative Numerical Study
of Premixed and Non-Premixed Ethylene Flames. Combustion Science
and Technology, 162(1), 19-35.
Egolfopoulos F. N. (1994). Geometric and radiation effects on steady and
unsteady strained laminar flames. Symposium (International) on
Combustion, 25(1), 1375-1381.
Egolfopoulos F. N., Cho P., and Law C. K. (1989). Laminar flame speeds of
methane-air mixtures under reduced and elevated pressures. Combustion
and Flame, 76(3–4), 375-391.
Egolfopoulos F. N., and Dimotakis P. E. (1998). Non-premixed hydrocarbon
ignition at high strain rates. Symposium (International) on Combustion,
27(1), 641-648.
Egolfopoulos F. N., and Dimotakis P. E. (2000). Effects of Additives on the Non-
Premixed Ignition of Ethylene in Air. Combustion Science and Technology,
156(1), 173-199.
Eisazadeh-Far K., Parsinejad F., and Metghalchi H. (2010). Flame structure and
laminar burning speeds of JP-8/air premixed mixtures at high
temperatures and pressures. Fuel, 89(5), 1041-1049.
El Bakali A., Braun-Unkhoff M., Dagaut P., Frank P., and Cathonnet M. (2000).
Detailed kinetic reaction mechanism for cyclohexane oxidation at pressure
up to ten atmospheres. Proceedings of the Combustion Institute, 28(2),
1631-1638.
Fieweger K., Blumenthal R., and Adomeit G. (1994). Shock-tube investigations
on the self-ignition of hydrocarbon-air mixtures at high pressures.
Symposium (International) on Combustion, 25(1), 1579-1585.
Fieweger K., Blumenthal R., and Adomeit G. (1997). Self-ignition of S.I. engine
model fuels: A shock tube investigation at high pressure. Combustion and
Flame, 109(4), 599-619.
Flowers D. L., Aceves S. M., Martinez-Frias J., and Dibble R. W. (2002).
Prediction of carbon monoxide and hydrocarbon emissions in iso-octane
HCCI engine combustion using multizone simulations. Proceedings of the
Combustion Institute, 29(1), 687-693.
190
Fotache C. G., Kreutz T. G., and Law C. K. (1997). Ignition of counterflowing
methane versus heated air under reduced and elevated pressures.
Combustion and Flame, 108(4), 442-470.
Fotache C. G., Kreutz T. G., and Law C. K. (1997). Ignition of hydrogen-enriched
methane by heated air. Combustion and Flame, 110(4), 429-440.
Fotache C. G., Kreutz T. G., Zhu D. L., and Law C. K. (1995). An experimental
study of ignition in nonpremixed counterflowing hydrogen versus heated
air. Combust. Sci. Technol., 109(1-6), 373-393.
Fotache C. G., Tan Y., Sung C. J., and Law C. K. (2000). Ignition of CO/H2/N2
versus heated air in counterflow: experimental and modeling results.
Combustion and Flame, 120(4), 417-426.
Fotache C. G., Wang H., and Law C. K. (1999). Ignition of ethane, propane, and
butane in counterflow jets of cold fuel versus hot air under variable
pressures. Combustion and Flame, 117(4), 777-794.
Franck J., Griffiths J. F., and Nimmo W. (1988). The control of spontaneous
ignition under rapid compression. Symposium (International) on
Combustion, 21(1), 447-454.
Gauthier B. M., Davidson D. F., and Hanson R. K. (2004). Shock tube
determination of ignition delay times in full-blend and surrogate fuel
mixtures. Combustion and Flame, 139(4), 300-311.
Gersen S., Mokhov A. V., Darmeveil J. H., and Levinsky H. B. (2010). Ignition
properties of n-butane and iso-butane in a rapid compression machine.
Combustion and Flame, 157(2), 240-245.
Glaude P. A., Warth V., Fournet R., Battin-Leclerc F., Scacchi G., and Côme G.
M. (1998). Modeling of the oxidation of n-octane and n-decane using an
automatic generation of mechanisms. International Journal of Chemical
Kinetics, 30(12), 949-959.
Gokulakrishnan P., Gaines G., Klassen M. S., and Roby R. J. (2007).
Autoignition of Aviation Fuels: Experimental and Modeling Study. Paper
presented at the 43rd AIAA/ASME/SAE/ASEE Joint Propulsion
Conference & Exhibit, 8-11 July, Cincinnati, OH.
191
Grana R., Seshadri K., Cuoci A., Niemann U., Faravelli T., and Ranzi E. (2012).
Kinetic modelling of extinction and autoignition of condensed hydrocarbon
fuels in non-premixed flows with comparison to experiment. Combustion
and Flame, 159(1), 130-141.
Granata S., Faravelli T., and Ranzi E. (2003). A wide range kinetic modeling
study of the pyrolysis and combustion of naphthenes. Combustion and
Flame, 132(3), 533-544.
Griffiths J. F., Halford-Maw P. A., and Rose D. J. (1993). Fundamental features
of hydrocarbon autoignition in a rapid compression machine. Combustion
and Flame, 95(3), 291-306.
Guéret C., Cathonnet M., Boettner J.-C., and Gaillard F. (1991). Experimental
study and modeling of kerosene oxidation in a jet-stirred flow reactor.
Symposium (International) on Combustion, 23(1), 211-216.
Halter F., Chauveau C., Djebaïli-Chaumeix N., and Gökalp I. (2005).
Characterization of the effects of pressure and hydrogen concentration on
laminar burning velocities of methane–hydrogen–air mixtures.
Proceedings of the Combustion Institute, 30(1), 201-208.
Haylett D. R., Cook R. D., Davidson D. F., and Hanson R. K. (2011). OH and
C2H4 species time-histories during hexadecane and diesel ignition behind
reflected shock waves. Proceedings of the Combustion Institute, 33(1),
167-173.
He X., Donovan M. T., Zigler B. T., Palmer T. R., Walton S. M., Wooldridge M. S.,
and Atreya A. (2005). An experimental and modeling study of iso-octane
ignition delay times under homogeneous charge compression ignition
conditions. Combustion and Flame, 142(3), 266-275.
He X., Zigler B. T., Walton S. M., Wooldridge M. S., and Atreya A. (2006). A rapid
compression facility study of OH time histories during iso-octane ignition.
Combustion and Flame, 145(3), 552-570.
Healy D., Donato N. S., Aul C. J., Petersen E. L., Zinner C. M., Bourque G., and
Curran H. J. (2010). Isobutane ignition delay time measurements at high
pressure and detailed chemical kinetic simulations. Combustion and
Flame, 157(8), 1540-1551.
192
Healy D., Donato N. S., Aul C. J., Petersen E. L., Zinner C. M., Bourque G., and
Curran H. J. (2010). n-Butane: Ignition delay measurements at high
pressure and detailed chemical kinetic simulations. Combustion and
Flame, 157(8), 1526-1539.
Healy D., Kalitan D. M., Aul C. J., Petersen E. L., Bourque G., and Curran H. J.
(2010). Oxidation of C1 −C5 Alkane Quinternary Natural Gas Mixtures at
High Pressures. Energy & Fuels, 24(3), 1521-1528.
Held T. J., Marchese A. J., and Dryer F. L. (1997). A semi-empirical reaction
mechanism for n-heptane oxidation and pyrolysis. Combustion Science
and Technology, 123(1-6), 107-146.
Herzler J., and Naumann C. (2008). Shock Tube Study of the Ignition of Lean
CO/H2 Fuel Blends at Intermediate Temperatures and High Pressure.
Combustion Science and Technology, 180(10-11), 2015-2028.
Hirasawa T., Sung C. J., Joshi A., Yang Z., Wang H., and Law C. K. (2002).
Determination of laminar flame speeds using digital particle image
velocimetry: Binary Fuel blends of ethylene, n-Butane, and toluene.
Proceedings of the Combustion Institute, 29(2), 1427-1434.
Holley A. T., Dong Y., Andac M. G., and Egolfopoulos F. N. (2006). Extinction of
premixed flames of practical liquid fuels: Experiments and simulations.
Combustion and Flame, 144(3), 448-460.
Holley A. T., Dong Y., Andac M. G., Egolfopoulos F. N., and Edwards T. (2007).
Ignition and extinction of non-premixed flames of single-component liquid
hydrocarbons, jet fuels, and their surrogates. Proceedings of the
Combustion Institute, 31(1), 1205-1213.
Holley A. T., You X. Q., Dames E., Wang H., and Egolfopoulos F. N. (2009).
Sensitivity of propagation and extinction of large hydrocarbon flames to
fuel diffusion. Proceedings of the Combustion Institute, 32(1), 1157-1163.
Hong Z., Lam K.-Y., Davidson D. F., and Hanson R. K. (2011). A comparative
study of the oxidation characteristics of cyclohexane, methylcyclohexane,
and n-butylcyclohexane at high temperatures. Combustion and Flame,
158(8), 1456-1468.
Horning D. C., Davidson D. F., and Hanson R. K. (2002). Study of the high-
temperature autoignition of n-alkane/O2/Ar mixtures. Journal of Propulsion
and Power, 18(2), 363-371.
193
Huber M. L., Smith B. L., Ott L. S., and Bruno T. J. (2008). Surrogate Mixture
Model for the Thermophysical Properties of Synthetic Aviation Fuel S-8:
Explicit Application of the Advanced Distillation Curve. Energy and Fuels,
22(2), 1104-1114.
Humer S., Frassoldati A., Granata S., Faravelli T., Ranzi E., Seiser R., and
Seshadri K. (2007). Experimental and kinetic modeling study of
combustion of JP-8, its surrogates and reference components in laminar
nonpremixed flows. Proceedings of the Combustion Institute, 31(1), 393-
400.
Humer S., Seiser R., and Seshadri K. (2002). Non-premixed and premixed
extinction and autoignition of C2H4, C2H6, C3H6, and C3H8. Proceedings
of the Combustion Institute, 29(2), 1597-1604.
Ji C., Dames E., Sirjean B., Wang H., and Egolfopoulos F. N. (2011). An
experimental and modeling study of the propagation of cyclohexane and
mono-alkylated cyclohexane flames. Proceedings of the Combustion
Institute, 33(1), 971-978.
Ji C., Dames E., Wang H., and Egolfopoulos F. N. (2012). Propagation and
extinction of benzene and alkylated benzene flames. Combustion and
Flame, 159(3), 1070-1081.
Ji C., Dames E., Wang Y. L., Wang H., and Egolfopoulos F. N. (2010).
Propagation and extinction of premixed C5–C12 n-alkane flames.
Combustion and Flame, 157(2), 277-287.
Ji C., and Egolfopoulos F. N. (2011). Flame propagation of mixtures of air with
binary liquid fuel mixtures. Proceedings of the Combustion Institute, 33(1),
955-961.
Ji C., Sarathy S. M., Veloo P. S., Westbrook C. K., and Egolfopoulos F. N. (2012).
Effects of fuel branching on the propagation of octane isomers flames.
Combustion and Flame, 159(4), 1426-1436.
Ji C., Wang Y. L., and Egolfopoulos F. N. (2011). Flame Studies of Conventional
and Alternative Jet Fuels. Journal of Propulsion and Power, 27(4), 856-
863.
Johnson M. V., Goldsborough S. S., Serinyel Z., O’Toole P., Larkin E., O’Malley
G., and Curran H. J. (2009). A Shock Tube Study of n- and iso-Propanol
Ignition. Energy & Fuels, 23(12), 5886-5898.
194
Jomaas G., Zheng X. L., Zhu D. L., and Law C. K. (2005). Experimental
determination of counterflow ignition temperatures and laminar flame
speeds of C2–C3 hydrocarbons at atmospheric and elevated pressures.
Proceedings of the Combustion Institute, 30(1), 193-200.
Kahandawala M. S. P., DeWitt M. J., Corporan E., and Sidhu S. S. (2008).
Ignition and Emission Characteristics of Surrogate and Practical Jet Fuels.
Energy & Fuels, 22(6), 3673-3679.
Kee R. J., Miller J. A., Evans G. H., and Dixon-Lewis G. (1989). A computational
model of the structure and extinction of strained, opposed flow, premixed
methane-air flames. Symposium (International) on Combustion, 22(1),
1479-1494.
Kee R. J., Rupley F. M., and Miller J. A. (1989). Chemkin-II: A FORTRAN
Chemical Kinetics Package for the Analysis of Gas-Phase Chemical
Kinetics, Sandia Report SAND89-8009: Sandia National Laboratories
Kee R. J., Warnatz J., and Miller J. A. (1983). A FORTRAN Computer Code
Package for the Evaluation of Gas-Phase Viscosities, Conductivities, and
Diffusion Coefficients, Sandia Report SAND83-8209: Sandia National
Laboratories
Kelley A. P., Liu W., Xin Y. X., Smallbone A. J., and Law C. K. (2011). Laminar
flame speeds, non-premixed stagnation ignition, and reduced mechanisms
in the oxidation of iso-octane. Proceedings of the Combustion Institute,
33(1), 501-508.
Kelly-Zion P. L., and Dec J. E. (2000). A computational study of the effect of fuel
type on ignition time in homogeneous charge compression ignition
engines Symposium (International) on Combustion, 28(1), 1187-1193.
KIM H., LIM Y., MIN K., and Lee D. (2002). Investigation of autoignition of
propane and n-butane blends using a rapid compression machine. KSME
International Journal, 16(8), 1127-1134.
Kojima S. (1994). Detailed modeling of n-butane autoignition chemistry.
Combustion and Flame, 99(1), 87-136.
Kojima S., and Suzuoki T. (1993). Autoignition-delay measurement over lean to
rich mixtures of n-butane/air under swirl conditions. Combustion and
Flame, 92(3), 254-265.
195
Kong S.-C., and Reitz R. D. (2003). Numerical study of premixed HCCI engine
combustion and its sensitivity to computational mesh and model
uncertainties. Combustion Theory and Modelling, 7(2), 417-433.
Kreutz T. G., and Law C. K. (1996). Ignition in nonpremixed counterflowing
hydrogen versus heated air: Computational study with detailed chemistry.
Combustion and Flame, 104(1–2), 157-175.
Kreutz T. G., and Law C. K. (1998). Ignition in Nonpremixed Counterflowing
Hydrogen versus Heated Air: Computational Study with Skeletal and
Reduced Chemistry. Combustion and Flame, 114(3–4), 436-456.
Kreutz T. G., Nishioka M., and Law C. K. (1994). The role of kinetic versus
thermal feedback in nonpremixed ignition of hydrogen versus heated air.
Combustion and Flame, 99(3–4), 758-766.
Kumar K., Mittal G., and Sung C.-J. (2009). Autoignition of n-decane under
elevated pressure and low-to-intermediate temperature conditions.
Combustion and Flame, 156(6), 1278-1288.
Kumar K., and Sung C.-J. (2010). A comparative experimental study of the
autoignition characteristics of alternative and conventional jet fuel/oxidizer
mixtures. Fuel, 89(10), 2853-2863.
Kumar K., and Sung C.-J. (2010). An experimental study of the autoignition
characteristics of conventional jet fuel/oxidizer mixtures: Jet-A and JP-8.
Combustion and Flame, 157(4), 676-685.
Kumar K., Sung C. J., and Hui X. (2009). Laminar Flame Speeds and Extinction
Limits of Conventional and Alternative Jet Fuels. Paper presented at the
47th AIAA Aerospace Sciences Meeting AIAA 2009-991, Orlando, Florida,
5–8 January.
Law C. K. (2006). Combustion Physics: Cambridge University Press, New York
(Chapter 8).
Lay T. H., Bozzelli J. W., Dean A. M., and Ritter E. R. (1995). Hydrogen Atom
Bond Increments for Calculation of Thermodynamic Properties of
Hydrocarbon Radical Species. The Journal of Physical Chemistry, 99(39),
14514-14527.
Lemaire O., Ribaucour M., Carlier M., and Minetti R. (2001). The production of
benzene in the low-temperature oxidation of cyclohexane, cyclohexene,
and cyclohexa-1,3-diene. Combustion and Flame, 127(1–2), 1971-1980.
196
Lenhert D. B., Miller D. L., and Cernansky N. P. (2007). THE OXIDATION OF JP-
8, JET-A, AND THEIR SURROGATES IN THE LOW AND
INTERMEDIATE TEMPERATURE REGIME AT ELEVATED
PRESSURES. Combustion Science and Technology, 179(5), 845-861.
Lindstedt R. P., and Maurice L. Q. (1995). DETAILED KINETIC MODELLING OF
N-HEPTANE COMBUSTION. Combustion Science and Technology,
107(4-6), 317-353.
Lindstedt R. P., and Maurice L. Q. (2000). Detailed Chemical-Kinetic Model for
Aviation Fuels. Journal of Propulsion and Power, 16(2), 187-195.
Liu N., Ji C., and Egolfopoulos F. N. Ignition of non-premixed cyclohexane and
mono-alkylated cyclohexane flames. Proceedings of the Combustion
Institute, 34 (2012) doi: 10.1016/j.proci.2012.05.104.
Liu N., Ji C., and Egolfopoulos F. N. (2012). Ignition of non-premixed C3–C12 n-
alkane flames. Combustion and Flame, 159(2), 465-475.
Liu N., Mani Sarathy S., Westbrook C. K., and Egolfopoulos F. N. Ignition of non-
premixed counterflow flames of octane and decane isomers. Proceedings
of the Combustion Institute 34 (2012),
http://dx.doi.org/10.1016/j.proci.2012.05.040.
Liu W., Kelley A. P., and Law C. K. (2010). Flame propagation and counterflow
nonpremixed ignition of mixtures of methane and ethylene. Combustion
and Flame, 157(5), 1027-1036.
Liu W., Zhu D., Wu N., and Law C. K. (2010). Ignition of n-heptane pool by
heated stagnating oxidizing flow. Combustion and Flame, 157(2), 259-266.
Lovell W. G. (1948). Knocking Characteristics of Hydrocarbons. Industrial &
Engineering Chemistry, 40(12), 2388-2438.
Marinov N. M., Pitz W. J., Westbrook C. K., Vincitore A. M., Castaldi M. J.,
Senkan S. M., and Melius C. F. (1998). Aromatic and Polycyclic Aromatic
Hydrocarbon Formation in a Laminar Premixed n-Butane Flame.
Combustion and Flame, 114(1–2), 192-213.
Maurice L. Q., Lander H., Edwards T., and Harrison Iii W. E. (2001). Advanced
aviation fuels: a look ahead via a historical perspective. Fuel, 80(5), 747-
756.
197
Mawid M. A., Park T. W., Sekar B., and Arana C. (2002). Development and
Validation of a Detailed JP-8 Fuel Chemistry Mechanism. Paper presented
at the 38th Joint Propulsion Conference and Exhibit AIAA 2002-3876,
Indianapolis, Indiana, 7-10 July.
Mehl M., Pitz W. J., Westbrook C. K., and Curran H. J. (2011). Kinetic modeling
of gasoline surrogate components and mixtures under engine conditions.
Proceedings of the Combustion Institute, 33(1), 193-200.
Middha P., and Wang H. (2005). First-principle calculation for the high-
temperature diffusion coefficients of small pairs: the H–Ar Case.
Combustion Theory and Modelling, 9(2), 353-363.
Minetti R., Carlier M., Ribaucour M., Therssen E., and Sochet L. R. (1995). A
rapid compression machine investigation of oxidation and auto-ignition of
n-Heptane: Measurements and modeling. Combustion and Flame, 102(3),
298-309.
Minetti R., Carlier M., Ribaucour M., Therssen E., and Sochet L. R. (1996).
Comparison of oxidation and autoignition of the two primary reference
fuels by rapid compression. Symposium (International) on Combustion,
26(1), 747-753.
Minetti R., Ribaucour M., Carlier M., Fittschen C., and Sochet L. R. (1994).
Experimental and modeling study of oxidation and autoignition of butane
at high pressure. Combustion and Flame, 96(3), 201-211.
Minetti R., Ribaucour M., Carlier M., and Sochet L. R. (1996). Autoignition Delays
of a Series of Linear and Branched Chain Alkanes in the Intermediate
Range of Temperature. Combustion Science and Technology, 113(1),
179-192.
Mittal G., and Sung C.-J. (2009). Autoignition of methylcyclohexane at elevated
pressures. Combustion and Flame, 156(9), 1852-1855.
Miyoshi A. (2010). Computational studies on the reactions of 3-butenyl and 3-
butenylperoxy radicals. International Journal of Chemical Kinetics, 42(5),
273-288.
Moc J., Black G., Simmie J. M., and Curran H. J. (2009). The Unimolecular
Decomposition and H Abstraction Reactions by HO and HO2 from n-
Butanol AIP Conf. Proc., 1148, 161-164.
198
Montgomery C. J., Cannon S. M., Mawid M. A., and Sekar B. (2002). Reduced
Chemical Kinetic Mechanisms for JP-8 Combustion. Paper presented at
the 40th AIAA Aerospace Sciences Meeting and Exhibit AIAA 2002-0036,
Reno, Nevada, 14-17 January.
Mourits F. M., and Rummens F. H. A. (1977). A critical evaluation of Lennard–
Jones and Stockmayer potential parameters and of some correlation
methods. Canadian Journal of Chemistry, 55(16), 3007-3020.
Mueller M. A., Yetter R. A., and Dryer F. L. (1999). Flow reactor studies and
kinetic modeling of the H2/O2/NOx and CO/H2O/O2/NOx reactions.
International Journal of Chemical Kinetics, 31(10), 705-724.
Natarajan J., Lieuwen T., and Seitzman J. (2007). Laminar flame speeds of
H2/CO mixtures: Effect of CO2 dilution, preheat temperature, and
pressure. Combustion and Flame, 151(1–2), 104-119.
Natarajan J., Nandula S., Lieuwen T., and Seitzman J. (2005). Laminar Flame
Speeds of Synthetic Gas Fuel Mixtures. ASME Conference Proceedings,
2005(4725X), 677-686.
Natelson R. H., Kurman M. S., Cernansky N. P., and Miller D. L. (2011). Low
temperature oxidation of n-butylcyclohexane. Combustion and Flame,
158(12), 2325-2337.
Nishioka M., Law C. K., and Takeno T. (1996). A flame-controlling continuation
method for generating S-curve responses with detailed chemistry.
Combustion and Flame, 104(3), 328-342.
Ó Conaire M., Curran H. J., Simmie J. M., Pitz W. J., and Westbrook C. K. (2004).
A comprehensive modeling study of hydrogen oxidation. International
Journal of Chemical Kinetics, 36(11), 603-622.
Oehlschlaeger M. A., Davidson D. F., Herbon J. T., and Hanson R. K. (2004).
Shock Tube Measurements of Branched Alkane Ignition Times and OH
Concentration Time Histories. International Journal of Chemical Kinetics,
36(2), 67-78.
Olchanski E., and Burcat A. (2006). Decane oxidation in a shock tube.
International Journal of Chemical Kinetics, 38(12), 703-713.
Orme J. P., Curran H. J., and Simmie J. M. (2005). Experimental and Modeling
Study of Methyl Cyclohexane Pyrolysis and Oxidation. The Journal of
Physical Chemistry A, 110(1), 114-131.
199
Park O., Veloo P. S., Liu N., and Egolfopoulos F. N. (2011). Combustion
characteristics of alternative gaseous fuels. Proceedings of the
Combustion Institute, 33(1), 887-894.
Petersen E. L., Kalitan D. M., Simmons S., Bourque G., Curran H. J., and
Simmie J. M. (2007). Methane/propane oxidation at high pressures:
Experimental and detailed chemical kinetic modeling. Proceedings of the
Combustion Institute, 31(1), 447-454.
Pitz W. J., Naik C. V., Mhaoldúin T. N., Westbrook C. K., Curran H. J., Orme J.
P., and Simmie J. M. (2007). Modeling and experimental investigation of
methylcyclohexane ignition in a rapid compression machine. Proceedings
of the Combustion Institute, 31(1), 267-275.
Puri P., Ma F., Choi J.-Y., and Yang V. (2005). Ignition characteristics of cracked
JP-7 fuel. Combustion and Flame, 142(4), 454-457.
Qin W., Egolfopoulos F. N., and Tsotsis T. T. (2001). Fundamental and
environmental aspects of landfill gas utilization for power generation.
Chemical Engineering Journal, 82(1–3), 157-172.
Ranzi E., Frassoldati A., Granata S., and Faravelli T. (2005). Wide-range kinetic
modeling study of the pyrolysis, partial oxidation, and combustion of heavy
n-alkanes. Industrial and Engineering Chemistry Research, 44(14), 5170-
5183.
Ranzi E., Gaffuri P., Faravelli T., and Dagaut P. (1995). A wide-range modeling
study of n-heptane oxidation. Combustion and Flame, 103(1–2), 91-106.
Ribaucour M., Minetti R., Sochet L. R., Curran H. J., Pitz W. J., and Westbrook C.
K. (2000). Ignition of isomers of pentane: An experimental and kinetic
modeling study. Proceedings of the Combustion Institute, 28(2), 1671-
1678.
Ristori A., Dagaut P., Bakali A. E., and Cathonnet M. (2001). The Oxidation of N-
Propylcyclohexane: Experimental Results and Kinetic Modeling.
Combustion Science and Technology, 165(1), 197-228.
Ritter E. R., and Bozzelli J. W. (1991). THERM: Thermodynamic property
estimation for gas phase radicals and molecules. International Journal of
Chemical Kinetics, 23(9), 767-778.
Rye L., and Wilson C. (2012). The influence of alternative fuel composition on
gas turbine ignition performance. Fuel, 96(0), 277-283.
200
Safta C., and Madnia C. K. (2006). Autoignition and structure of nonpremixed
CH4/H2 flames: Detailed and reduced kinetic models. Combustion and
Flame, 144(1–2), 64-73.
Salooja K. C. (1962). Studies of combustion processes leading to ignition of
isomeric hexanes. Combustion and Flame, 6(0), 275-285.
Salooja K. C. (1965). The degenerate chain braching intermediate in
hydrocarbon combustion: Some evidence from studies on the isomeric
hexanes. Combustion and Flame, 9(3), 219-227.
Sarathy S. M., Niemann U., Yeung C., Gehmlich R., Westbrook C. K., Plomer M.,
Luo Z., Mehl M., Pitz W. J., Seshadri K., Thomson M. J., and Lu T. A
counterflow diffusion flame study of branched octane isomers.
Proceedings of the Combustion Institute 34 (2012),
http://dx.doi.org/10.1016/j.proci.2012.05.106.
Sarathy S. M., Westbrook C. K., Mehl M., Pitz W. J., Togbe C., Dagaut P., Wang
H., Oehlschlaeger M. A., Niemann U., Seshadri K., Veloo P. S., Ji C.,
Egolfopoulos F. N., and Lu T. (2011). Comprehensive chemical kinetic
modeling of the oxidation of 2-methylalkanes from C7 to C20. Combustion
and Flame, 158(12), 2338-2357.
Sarathy S. M., Yeung C., Westbrook C. K., Pitz W. J., Mehl M., and Thomson M.
J. (2011). An experimental and kinetic modeling study of n-octane and 2-
methylheptane in an opposed-flow diffusion flame. Combustion and Flame,
158(7), 1277-1287.
Saxena P., and Williams F. A. (2006). Testing a small detailed chemical-kinetic
mechanism for the combustion of hydrogen and carbon monoxide.
Combustion and Flame, 145(1–2), 316-323.
Seiser R., Pitsch H., Seshadri K., Pitz W. J., and Gurran H. J. (2000). Extinction
and autoignition of n-heptane in counterflow configuration. Proceedings of
the Combustion Institute, 28(2), 2029-2037.
Seshadri K., Humer S., and Seiser R. (2008). Activation-energy asymptotic
theory of autoignition of condensed hydrocarbon fuels in non-premixed
flows with comparison to experiment. Combustion Theory and Modelling,
12(5), 831-855.
201
Silke E. J., Curran H. J., and Simmie J. M. (2005). The influence of fuel structure
on combustion as demonstrated by the isomers of heptane: a rapid
compression machine study. Proceedings of the Combustion Institute,
30(2), 2639-2647.
Silke E. J., Pitz W. J., Westbrook C. K., and Ribaucour M. (2007). Detailed
Chemical Kinetic Modeling of Cyclohexane Oxidation†. The Journal of
Physical Chemistry A, 111(19), 3761-3775.
Sirjean B., Buda F., Hakka H., Glaude P. A., Fournet R., Warth V., Battin-Leclerc
F., and Ruiz-Lopez M. (2007). The autoignition of cyclopentane and
cyclohexane in a shock tube. Proceedings of the Combustion Institute,
31(1), 277-284.
Sirjean B., Dames E., Sheen D. A., You X.-Q., Sung C., Holley A. T.,
Egolfopoulos F. N., Wang H., Vasu S. S., Davidson D. F., Hanson R. K.,
Pitsch H., Bowman C. T., Kelley A., Law C. K., Tsang W., Cernansky N. P.,
Miller D. L., Violi A., and Lindstedt R. P. (September 15, 2009). A high-
temperature chemical kinetic model of n-alkane oxidation, JetSurF version
1.0, (http://melchior.usc.edu/JetSurF/Version1_0/Index.html).
Smallbone A. J., Liu W., Law C. K., You X. Q., and Wang H. (2009).
Experimental and modeling study of laminar flame speed and non-
premixed counterflow ignition of n-heptane. Proceedings of the
Combustion Institute, 32(1), 1245-1252.
Tanaka S., Ayala F., Keck J. C., and Heywood J. B. (2003). Two-stage ignition in
HCCI combustion and HCCI control by fuels and additives. Combustion
and Flame, 132(1–2), 219-239.
Tang C., He J., Huang Z., Jin C., Wang J., Wang X., and Miao H. (2008).
Measurements of laminar burning velocities and Markstein lengths of
propane–hydrogen–air mixtures at elevated pressures and temperatures.
International Journal of Hydrogen Energy, 33(23), 7274-7285.
Trujillo J. Y. D., Kreutz T. G., and Law C. K. (1997). Ignition in a Counterflowing
Non-Premixed CO/H2-Air System. Combustion Science and Technology,
127(1-6), 1-27.
Vanderover J., and Oehlschlaeger M. A. (2009). Ignition time measurements for
methylcylcohexane- and ethylcyclohexane-air mixtures at elevated
pressures. International Journal of Chemical Kinetics, 41(2), 82-91.
202
Varatharajan B., and Williams F. A. (2002). Ethylene Ignition and Detonation
Chemistry, Part 1: Detailed Modeling and Experimental Comparison.
Journal of Propulsion and Power, 18(2), 344-351.
Vasu S. S., Davidson D. F., and Hanson R. K. (2008). Jet fuel ignition delay
times: Shock tube experiments over wide conditions and surrogate model
predictions. Combustion and Flame, 152(1–2), 125-143.
Vasu S. S., Davidson D. F., and Hanson R. K. (2009). OH time-histories during
oxidation of n-heptane and methylcyclohexane at high pressures and
temperatures. Combustion and Flame, 156(4), 736-749.
Vasu S. S., Davidson D. F., Hong Z., and Hanson R. K. (2008). Shock Tube
Study of Methylcyclohexane Ignition over a Wide Range of Pressure and
Temperature. Energy & Fuels, 23(1), 175-185.
Vasu S. S., Davidson D. F., Hong Z., Vasudevan V., and Hanson R. K. (2009). n-
Dodecane oxidation at high-pressures: Measurements of ignition delay
times and OH concentration time-histories. Proceedings of the
Combustion Institute, 32(1), 173-180.
Vermeer D. J., Meyer J. W., and Oppenheim A. K. (1972). Auto-ignition of
hydrocarbons behind reflected shock waves. Combustion and Flame,
18(3), 327-336.
Violi A., Yan S., Eddings E. G., Sarofim A. F., Granata S., Faravelli T., and Ranzi
E. (2002). Experimental formulation and kinetic model for JP-8 surrogate
mixtures. Combustion Science and Technology, 174(11-12), 399-417.
Vlachos D. G., Schmidt L. D., and Aris R. (1993). Ignition and extinction of flames
near surfaces: Combustion of H2 in air. Combustion and Flame, 95(3),
313-335.
Voisin D., Marchal A., Reuillon M., Boettner J. C., and Cathonnet M. (1998).
Experimental and Kinetic Modeling Study of Cyclohexane Oxidation in a
JSR at High Pressure. Combustion Science and Technology, 138(1-6),
137-158.
Wang H., Dames E., Sirjean B., and al. e. (September 19, 2010). A High-
Temperature Chemical Kinetic Model of n-Alkane (up to n-Dodecane),
Cyclohexane, and Methyl-, Ethyl-, n-Propyl and n-Butyl-Cyclohexane
Oxidation at High Temperatures (JetSurF Version 2.0).
203
Wang H., and Oehlschlaeger M. A. (2012). Autoignition studies of conventional
and Fischer–Tropsch jet fuels. Fuel, 98(0), 249-258.
Wang H., You X., Joshi A. V., Davis S. G., Laskin A., Egolfopoulos F. N., and
Law C. K. (May 2007). High-Temperature Combustion Reaction Model of
H2/CO/C1–C4 Compounds (USC Mech Version II).
Wang Y. L., Feng Q., Egolfopoulos F. N., and Tsotsis T. T. (2011). Studies of C4
and C10 methyl ester flames. Combustion and Flame, 158(8), 1507-1519.
Westbrook C. K. (2000). Chemical kinetics of hydrocarbon ignition in practical
combustion systems. Proceedings of the Combustion Institute, 28(2),
1563-1577.
Westbrook C. K., Pitz W. J., Boercker J. E., Curran H. J., Griffiths J. F.,
Mohamed C., and Ribaucour M. (2002). Detailed chemical kinetic reaction
mechanisms for autoignition of isomers of heptane under rapid
compression. Proceedings of the Combustion Institute, 29(1), 1311-1318.
Westbrook C. K., Pitz W. J., Herbinet O., Curran H. J., and Silke E. J. (2009). A
comprehensive detailed chemical kinetic reaction mechanism for
combustion of n-alkane hydrocarbons from n-octane to n-hexadecane.
Combustion and Flame, 156(1), 181-199.
Wu C. K., and Law C. K. (1985). On the determination of laminar flame speeds
from stretched flames. Symposium (International) on Combustion, 20(1),
1941-1949.
Yang Y., and Boehman A. L. (2009). Experimental study of cyclohexane and
methylcyclohexane oxidation at low to intermediate temperature in a
motored engine. Proceedings of the Combustion Institute, 32(1), 419-426.
You X., Egolfopoulos F. N., and Wang H. (2009). Detailed and simplified kinetic
models of n-dodecane oxidation: The role of fuel cracking in aliphatic
hydrocarbon combustion. Proceedings of the Combustion Institute, 32(1),
403-410.
Yu G., Law C. K., and Wu C. K. (1986). Laminar flame speeds of hydrocarbon +
air mixtures with hydrogen addition. Combustion and Flame, 63(3), 339-
347.
204
Zeppieri S., Brezinsky K., and Glassman I. (1997). Pyrolysis studies of
methylcyclohexane and oxidation studies of methylcyclohexane and
methylcyclohexane/toluene blends. Combustion and Flame, 108(3), 266-
286.
Zeppieri S. P., Klotz S. D., and Dryer F. L. (2000). Modeling concepts for larger
carbon number alkanes: A partially reduced skeletal mechanism for n-
decane oxidation and pyrolysis. Proceedings of the Combustion Institute,
28(2), 1587-1595.
Zhang H. R., Eddings E. G., and Sarofim A. F. (2007). Criteria for selection of
components for surrogates of natural gas and transportation fuels.
Proceedings of the Combustion Institute, 31(1), 401-409.
Zheng X. L., Blouch J. D., Zhu D. L., Kreutz T. G., and Law C. K. (2002). Ignition
of premixed hydrogen/air by heated counterflow. Proceedings of the
Combustion Institute, 29(2), 1637-1643.
Zheng X. L., and Law C. K. (2004). Ignition of premixed hydrogen/air by heated
counterflow under reduced and elevated pressures. Combustion and
Flame, 136(1–2), 168-179.
Zhukov V. P., Sechenov V. A., and Starikovskii A. Y. (2008). Autoignition of n-
decane at high pressure. Combustion and Flame, 153(1–2), 130-136.
Abstract (if available)
Abstract
Practical jet fuels are widely used in air-breathing propulsion, but the chemical mechanisms that control their combustion are not yet understood. Thousands of components are contained in conventional and alternative jet fuels, making thus any effort to model their combustion behavior a daunting task. That has been the motivation behind the development of surrogate fuels that contain typically a small number of neat components, whose physical properties and combustion behavior mimic those of the real jet fuel, and whose kinetics could be modeled with increased degree of confidence. Towards that end, a large number of experimental data are required both for the real fuels and the attendant surrogate components that could be used to develop and validate detailed kinetic models. Those kinetic models could be used then upon reduction to model a combustor and eventually optimize its performance. ❧ Among all flame phenomena, ignition is rather sensitive to the oxidative and pyrolytic propensity of the fuel as well as to its diffusivity. The counterflow configuration is ideal in probing both the fuel reactivity and diffusivity aspects of the ignition process and it was used in the present work to determine the ignition temperatures of premixed and non-premixed flames of a variety of fuels relevant to air-breathing propulsion. The experiments were performed at atmospheric pressure, elevated unburned fuel mixture temperatures, and various strain rates that were measured locally. Several recent kinetic models were used in direct numerical simulations of the experiments and the computed results were tested against the experimental data. Furthermore, through sensitivity, reaction path, and structure analyses of the computed flames, insight was provided into the dominant mechanisms that control ignition. It was found that ignition is primarily sensitive to fuel diffusion and secondarily sensitive to chemical kinetics and intermediate species diffusivities under the low fuel concentrations. As for the detailed high temperature oxidation chemistry, ignition of normal, branched, and cyclic alkane flames were found to be sensitive largely to H₂/CO and C₁-C₄ small hydrocarbon chemistry, while for branched alkanes fuel-related reactions do have accountable effect on ignition due to the low rate of initial fuel decomposition that limits the overall reactions preceding ignition. Analyses of the computed flame structures revealed that the concentrations of ignition-promoting radicals such as H, HCO, C₂H₃, and OH, and ignition-inhibiting radicals such as C₃H₆, aC₃H₅, and CH₃ are key to the occurrence of ignition. ❧ Finally, the ignition characteristics of conventional and alternative jet fuels were studied and were to correlate with the chemical classifications and diffusivities of the neat species that are present in the practical fuel.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Studies on the flame dynamics and kinetics of alcohols and liquid hydrocarbon fuels
PDF
Modeling investigations of fundamental combustion phenomena in low-dimensional configurations
PDF
Investigations of fuel effects on turbulent premixed jet flames
PDF
CFD design of jet-stirred chambers for turbulent flame and chemical kinetics experiments
PDF
Experimental and kinetic modeling studies of flames of H₂, CO, and C₁-C₄ hydrocarbons
PDF
A theoretical study of normal alkane combustion
PDF
Pressure effects on C₁-C₂ hydrocarbon laminar flames
PDF
Accuracy and feasibility of combustion studies under engine relevant conditions
PDF
Studies of methane counterflow flames at low pressures
PDF
Studies of combustion characteristics of heavy hydrocarbons in simple and complex flows
PDF
Catalytic methane ignition over freely-suspended palladium nanoparticles
PDF
Kinetic modeling of high-temperature oxidation and pyrolysis of one-ringed aromatic and alkane compounds
PDF
Studies of siloxane decomposition in biomethane combustion
PDF
Experimental studies of high pressure combustion using spherically expanding flames
PDF
Experimental investigation of the propagation and extinction of edge-flames
PDF
Investigations of fuel and hydrodynamic effects in highly turbulent premixed jet flames
PDF
Mesoscale SOFC-based power generator system: modeling and experiments
PDF
Flame characteristics in quasi-2D channels: stability, rates and scaling
PDF
End-gas autoignition investigations using confined spherically expanding flames
PDF
Re-assessing local structures of turbulent flames via vortex-flame interaction
Asset Metadata
Creator
Liu, Ning
(author)
Core Title
Flame ignition studies of conventional and alternative jet fuels and surrogate components
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Mechanical Engineering
Publication Date
01/17/2013
Defense Date
11/06/2012
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
flame ignition,jet fuels,kinetic modeling,OAI-PMH Harvest,surrogate components
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Egolfopoulos, Fokion N. (
committee chair
), Ronney, Paul D. (
committee member
), Tsotsis, Theodore T. (
committee member
)
Creator Email
liuning29@gmail.com,liuning29@live.cn
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-128666
Unique identifier
UC11292118
Identifier
usctheses-c3-128666 (legacy record id)
Legacy Identifier
etd-LiuNing-1393.pdf
Dmrecord
128666
Document Type
Dissertation
Rights
Liu, Ning
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
flame ignition
jet fuels
kinetic modeling
surrogate components